Sexual Assault Defence in Alberta

Anyone charged with a sexual offence in Alberta is reading about a body of law that has changed significantly since 2018 and continues to evolve. Parliament has amended the consent provisions and the rules governing the use of the complainant's records at trial. The Supreme Court of Canada has decided major cases on consent, on the mistaken-belief defence, on disclosure, and on the pre-trial applications the defence has to bring before certain evidence can be used. Another important Supreme Court decision on the mental element of the offence was argued in May 2026 and is currently under reserve. The framework on this page reflects the law as it stands today, with that pending decision flagged where it bears on the analysis.

This page is a working overview of how a sexual assault prosecution is structured in Alberta — what the Crown must prove, how consent and mistaken belief operate, how credibility is analyzed at trial, how identification is handled, how disclosure works on both sides, what applications the defence has to bring before trial, and what a conviction looks like at sentencing. The page is not legal advice. Anyone facing a charge should speak with counsel early; many of the decisions on a sexual offence file have to be made well before the trial date.

What the Crown must prove

A sexual assault prosecution under section 271 of the Criminal Code requires the Crown to prove three things about what was done, and two things about what the accused was thinking when he did it. The architecture is the same as any other criminal offence — actus reus and mens rea — but the elements have specific content shaped by decades of Supreme Court of Canada jurisprudence on this offence.

On the actus reus, the Crown has to prove that the accused intentionally applied force to the complainant, that the force was sexual in nature, and that it was applied without the complainant's consent. The Supreme Court of Canada in R v Chase framed the sexual-nature question as an objective inquiry: viewed in the light of all the circumstances, would a reasonable observer regard the conduct as sexual? The part of the body touched, the nature of the contact, the situation, the words and gestures accompanying the act, and the apparent purpose of the person doing the touching all bear on that question. Sexual gratification is not required; what is required is that the conduct, viewed objectively, was sexual. The Supreme Court of Canada in R v Ewanchuk confirmed that the third element — absence of consent — turns on the complainant's actual state of mind at the time, not on what her conduct may have suggested to others. That element is the subject of Section 2.

On the mens rea, the Crown has to prove that the accused intended to touch the complainant and that he knew the complainant had not consented, or was reckless or wilfully blind to the absence of consent. The defence of honest but mistaken belief in communicated consent, where it arises, operates as a denial of that second element. Section 2 of this page works through the consent question and the mistaken-belief defence in detail, along with a current Supreme Court of Canada appeal that may refine what the Crown must prove on the mens rea side when mistaken belief is not in play.

Sections 272 and 273 of the Criminal Code create the more serious offences of sexual assault with a weapon or causing bodily harm, and aggravated sexual assault. Each requires the Crown to prove additional elements beyond those of the simple offence under section 271 — the use of a weapon, the causing of bodily harm, or conduct that wounds, maims, disfigures, or endangers the life of the complainant. The architecture is otherwise the same. Section 7 of this page discusses what a conviction for each of these offences would mean at sentencing.

Consent and mistaken belief in communicated consent

Consent and mistaken belief in consent are not two separate defences. They are two halves of one analysis. Consent asks whether the sexual activity happened without the complainant's agreement. Mistaken belief asks whether, even if there was no agreement, the accused honestly believed there was — and took the steps the law required him to take to be sure. The two questions run in sequence. Treating them as alternatives is a common error in this area of law.

Consent: did the complainant actually agree?

Section 273.1 of the Criminal Code defines consent as the complainant's voluntary agreement to the sexual activity in question. The Supreme Court of Canada settled the meaning in R v Ewanchuk: consent means what the complainant actually wanted at the time. It is the complainant's internal state of mind that matters at this stage, not what the surrounding circumstances suggest. Consent cannot be inferred from silence, from ambiguous conduct, or from what happened on an earlier occasion.

This does not mean the complainant's evidence about her state of mind decides the case on its own. The complainant's testimony is the only direct evidence of what she was thinking, but the trial judge must still assess that testimony against the totality of the evidence, including any conduct that may appear inconsistent with her account. Ewanchuk makes the structure clear. The complainant's words and actions before and during the encounter can raise a reasonable doubt about her assertion that she did not want the sexual touching to take place. If the trial judge accepts the complainant's evidence that she did not consent, the Crown has proven the absence of consent — no matter how strongly the complainant's conduct might cut against that finding. But the judge does not have to accept it. The trier of fact can reach only one of two conclusions: the complainant consented or she did not. There is no third option. Which conclusion follows depends on what the judge or jury, weighing all the evidence, is prepared to find about the complainant's state of mind at the time.

The Criminal Code also identifies situations in which the law treats consent as not having been given, regardless of what the complainant felt or said in the moment. These include consent given by someone other than the complainant, consent given by a person who could not consent, consent obtained through an abuse of trust or authority, and — added by Parliament in 2018 — consent purportedly given while the complainant was unconscious. The Supreme Court in R v JA had already ruled that a person cannot consent in advance to sexual activity that will happen during a period of unconsciousness. The 2018 amendment put that ruling into the statute.

Intoxication raises a separate question. A complainant who is intoxicated may or may not have the capacity to consent. The Supreme Court in R v GF set out the framework. The trial judge asks whether the complainant could still understand the sexual nature of the activity, recognize the person they were with, and know they had the option to refuse. Capacity and consent are separate findings. Even where a complainant had the capacity to consent, the Crown still has to prove that consent was actually given.

Fear, threats, and the exercise of authority can also remove consent. Section 265(3) of the Code deems an absence of consent where the complainant submits because force was applied, because she feared force would be applied, because of fraud, or because authority was used over her. Ewanchuk read that section as a subjective inquiry. If the complainant honestly believed she had to comply to avoid being harmed, the law treats that as no consent — and her fear does not have to be objectively reasonable, or even communicated to the accused, for the section to apply.

Deception can also vitiate consent. In R v Hutchinson, the Supreme Court considered a case where the complainant had agreed to sexual activity on the condition that the accused use a condom, and the accused had sabotaged the condom without her knowledge. The Court structured the analysis in two steps. First, was there voluntary agreement to the sexual activity itself — meaning the specific physical sex act, its sexual nature, and the specific identity of the partner (defined narrowly as the particular person the complainant believed she was with)? Second, even if there was, was that agreement undermined by fraud or by one of the other circumstances the Code recognizes as vitiating consent? Hutchinson placed conditions like STI status outside the first step and inside the second: they are not part of defining the "sexual activity in question" itself, but a sufficiently deceptive breach of a condition the complainant insisted on can vitiate what otherwise looked like agreement. The two-step structure still governs any case where deception is alleged to have undermined consent, but the Supreme Court has since refined how it applies to condom use. In R v Kirkpatrick, the Court held by a narrow majority that where the complainant has agreed to sex on the condition that the accused wear a condom, and the accused does not, the case is decided at the first step — sex with a condom and sex without one are different physical acts, and consent to one is not consent to the other. Hutchinson still governs the narrower situation it actually decided: deceptive sabotage of an agreed-upon condom, which goes to step two as fraud.

Mistaken belief: did the accused honestly think the complainant was consenting?

If the Crown proves the sexual activity happened without consent in the sense described above, the accused is not automatically guilty. There is one remaining avenue: the defence of honest but mistaken belief in communicated consent under section 273.2 of the Criminal Code. The Supreme Court in R v Barton sharpened how the defence is even named. It is not simply a belief in "consent" — it is a belief that the complainant communicated consent, by words or by active conduct. That refinement matters, because a belief in what the complainant was actually thinking, without something she said or did to communicate it, is no defence at all.

The Code blocks the defence in several situations. It is not available where the accused's belief came from being intoxicated, from being reckless, or from being wilfully blind to whether the complainant was consenting. It is not available unless the accused took steps that, in the circumstances he knew at the time, were reasonable to ascertain that the complainant was consenting. And — since the 2018 amendments — it is not available unless there is evidence the complainant communicated consent by words or active conduct. Passive non-resistance is not enough.

The reasonable-steps requirement, in Barton's words, is a precondition: no reasonable steps, no defence. The requirement has both an objective and a subjective dimension. The accused must have taken steps that were objectively reasonable, and the reasonableness of those steps is assessed in light of the circumstances the accused actually knew at the time. The Supreme Court has consistently declined to reduce the requirement to a checklist. What was said before the activity, the nature of the relationship, any signs of intoxication or distress, and the specific activity in question all bear on the analysis. Consent to one kind of sexual activity is not consent to another. The steps the accused took have to address the activity that actually occurred. And steps grounded in stereotypes about women or in mistakes of law about what consent means cannot count as reasonable.

The defence cannot even reach a jury unless there is a sufficient evidentiary foundation for it. Before it is put, the trial judge has to be satisfied that there is some evidence both that the accused took reasonable steps and that he honestly believed the complainant communicated consent. If the evidence on either branch is absent, the defence has no air of reality and must not be left with the jury. If the defence does have an air of reality, the Crown can defeat it by negating either branch beyond a reasonable doubt — by proving that the accused did not take steps that were reasonable in the circumstances known to him at the time, or by proving that the accused did not actually hold an honest belief that consent had been communicated. The two requirements are not interchangeable. Reasonable steps is the statutory precondition that gates access to the defence; honest belief in communicated consent is the substantive mens rea question the defence is built to answer. Even where the Crown cannot displace the steps, the defence fails if the honest belief is proven false beyond a reasonable doubt — and even where the Crown cannot displace the honest belief, the defence fails if the steps were not taken.

A note on a pending Supreme Court appeal

One important caveat affects the framework above as it stands today. The Supreme Court of Canada is considering an appeal that touches on the mental element required of the Crown in a sexual assault prosecution — specifically, the part of the analysis that comes after the question of mistaken belief in communicated consent has been answered against the accused, or that arises in cases where the defence of mistaken belief is not available at all.

The case is R v Bilinski, 2025 ABCA 270. The Alberta Court of Appeal considered a Crown appeal from a jury acquittal on a sexual assault charge where the accused had not pursued the defence of honest but mistaken belief in communicated consent at trial. The question on appeal was a narrow but important one: what does the Crown still have to prove about the accused's state of mind, and how must a jury be instructed, when the defence of mistaken belief is not in play? The Alberta court held that the Crown still bears the burden of proving, beyond a reasonable doubt, that the accused knew the complainant had not communicated consent, or was reckless or wilfully blind to the absence of communicated consent. It also held that in cases where the mistaken-belief defence has not been advanced or has no air of reality, the trial judge should give what has come to be called the "little difficulty" instruction: that if the Crown has proven the complainant did not consent, the jury should have little difficulty concluding that the accused knew, or was reckless or wilfully blind to, the absence of consent. The court was unanimous that the trial judge's failure to give that instruction was an error of law, though the panel divided on whether the error was material enough to set the acquittal aside. The Alberta Court of Appeal has since applied that reasoning in R v Hutton, where a unanimous panel set aside a sexual assault conviction and entered an acquittal on the basis that the trial judge had not addressed the Crown's burden of proving knowledge, recklessness, or wilful blindness as to the absence of consent — a question that had to be answered even where the defence of mistaken belief was found unavailable.

The Supreme Court of Canada granted Mr. Bilinski leave to appeal on February 9, 2026, limited by agreement of the parties to a single question: what mental element must the Crown prove in a sexual assault prosecution where the defence of honest but mistaken belief in communicated consent does not arise, and how should juries be instructed to assess that mental element? The Court heard the appeal on May 22, 2026, alongside a companion appeal from the Ontario Court of Appeal's decision in R v Degale, which reached substantively the same conclusion on the mens rea question. The Supreme Court had reserved judgment in both appeals when this page was last updated. The eventual decisions will likely refine, and may significantly shift, what the Crown must prove on the mens rea side of the offence in every sexual assault prosecution — not only those where mistaken belief is in play. Anyone consulting this page in connection with a current matter should check the cases before relying on the framework above.

What this means in practical terms

Three things follow from the architecture above for someone navigating a sexual assault charge.

The first is that the two inquiries run on different evidence. Whether the complainant consented turns on the complainant's testimony about what she was thinking, assessed by the trial judge against the totality of the evidence — including conduct that may cut against her account. Whether the accused honestly believed that consent had been communicated turns on the accused's testimony about what he understood her words and conduct to convey, what he did in the moment to make sure, and what was happening around him. These are not interchangeable lines of defence. They can be advanced together, but they are not the same argument.

The second is that the mistaken-belief defence requires evidence — typically from the accused — about what was actually done in the moment to confirm consent. Without that evidence, the defence will not clear the threshold the trial judge has to apply before the defence can even be put to a jury. Whether the defence is viable on a given set of facts depends on what that evidence will look like, and that question is best addressed with counsel early.

The third is that the law on consent and mistaken belief has changed significantly since 2018, and continues to evolve. Earlier case law on what the defence required — particularly on whether passive non-resistance could ground a belief in consent — does not survive the current statutory framework intact. The Supreme Court's pending decision in Bilinski will likely sharpen the picture further on the mens rea question. Anyone advising on a current matter, or trying to understand how the law applies to events from before 2018, needs to be aware of which version of the law governs.

A credibility-based path to acquittal

A sexual assault trial in which two people give different accounts of the same event is, in the popular imagination, decided by which person the trier of fact believes. Canadian law has not framed the question that way for more than thirty years. A finding that the accused is not believed is not a finding of guilt. A finding that the complainant is believed is not, on its own, a finding of guilt either. Between the two accounts lies a path to acquittal that does not require the trier of fact to choose one as the truth.

The framework: W(D) and the question that controls

The Supreme Court of Canada in R v W(D) articulated a three-step framework for cases that turn on competing testimony. First, if the trier of fact believes the accused's evidence, the accused must be acquitted. Second, even if the trier does not believe the accused's evidence but is left in reasonable doubt by it, the accused must be acquitted. Third, even if the trier is not left in doubt by the accused's evidence, the trier must still decide, on the basis of the evidence it does accept, whether the Crown has proven guilt beyond a reasonable doubt.

The framework is not a formula. In R v Vuradin, the Supreme Court held that a trial judge is not required to recite the steps; what matters is whether the principles have been applied. A verdict of guilt must not be based on a choice between the accused's evidence and the Crown's evidence.

Credibility and reliability are not the same question. The Ontario Court of Appeal in R v Morrissey drew the distinction directly. Credibility concerns a witness's sincerity — her willingness to speak the truth as she believes it to be. Reliability concerns the accuracy of her testimony — her ability to observe, recall, and recount events accurately. A credible witness can give unreliable evidence. An honest witness whose perception, memory, or recall is unreliable on a point can leave the trier of fact in reasonable doubt even where there is no question of dishonesty. The distinction matters because credibility-driven trials are rarely decided on the basis that someone is lying. They are more often decided on the basis that an honest witness, doing her best, has produced evidence the trier is not prepared to act on with the certainty the criminal standard demands.

The Alberta Court of Appeal in R v Ryon sharpened how the framework operates in modern sexual offence trials. The protection is not confined to the accused's own testimony — it applies to all of the exculpatory evidence, whether tendered by the Crown or the accused, and whether or not the accused has testified at all. The framework also does not require the trial judge to consider the accused's evidence first, or in any particular order. The court was equally clear about what the framework does not do. Even where the trier rejects the accused's evidence entirely, mere rejection is not, on its own, proof that the Crown's account is true. The trier has to return to the evidence it does accept and ask whether that evidence proves the offence beyond a reasonable doubt.

The same court has also been candid about the limits of that reframing. Later in 2019, in R v Achuil, Justice Martin — who had written for the court in Ryon — qualified the language he had proposed. The phrase he had used in Ryon to capture the second step of the analysis, that the trier should acquit if "not confident they can accept the Crown's version of events," was not a recognized legal standard, and instructing a jury in those terms would be an error. The safer formulation, the court held, is the older one: that if the accused's evidence is believed, or even if not believed still leaves the trier with a reasonable doubt that it may be true, the trier must acquit. The court in Achuil also held that the W(D) analysis does not engage at all in a particular kind of case — one where the accused's own evidence, even if accepted in full, establishes guilt beyond a reasonable doubt. There, the framework adds nothing because the question it is designed to answer does not arise. Justice Slatter, concurring in the result, added a point that bears on every credibility-driven trial: the trial judge's primary task is to make findings of fact, not to summarize the evidence; reciting what the witnesses said is not the same thing as deciding what happened, and a trial judge cannot sit on the fence between competing accounts.

The standard the framework operates against is not "more likely than not." In R v Lifchus, the Supreme Court explained that a reasonable doubt is a doubt grounded in reason and common sense, logically connected to the evidence or to the absence of evidence. In R v Starr, the Court placed the standard on the spectrum of proof: much closer to absolute certainty than to a balance of probabilities. Less than absolute certainty is required; more than probable guilt is required.

The second step of W(D) does most of the work in practice. The Supreme Court returned to it in 2024 in R v Kruk. Justice Martin restated the principle in plain modern terms: an accused cannot be found guilty simply because he is disbelieved. Some elements of the totality of the evidence may give rise to a reasonable doubt even where much — or all — of the accused's evidence is disbelieved. Where the trier of fact does not know whether to believe the accused, or does not know who to believe, the accused is entitled to an acquittal.

The practical consequence is that a trial that produces evidence the trier does not fully accept — but that also does not produce the Crown evidence the trier is prepared to act on with certainty — can result in acquittal even though the accused was not, on the trial judge's view, telling the whole truth. That is not a defence-side aspiration. It is the operative consequence of the Supreme Court's reasoning in W(D) as restated in Kruk.

The right to silence, in fact and in law

An accused is not required to testify, and the trier of fact cannot treat a decision not to testify as evidence of guilt. The Supreme Court in Kruk set out the architecture: the presumption of innocence and the right to silence work together to restrict how a trial judge may reason about credibility. A trial judge may not discount the credibility of the accused on the basis that the accused, facing a criminal charge, has reason to lie. The protection has two faces. The de jure rule is that silence cannot be evidence of guilt. The de facto rule follows from W(D) step two as restated in Kruk — an accused who chooses to testify and is disbelieved has not, by failing as a witness, proved the Crown's case.

The framework also operates upstream of the question discussed in Section 1 about what the Crown must prove on the mental element of the offence — a question the Supreme Court's pending decision in R v Bilinski will address. The credibility analysis described here is logically prior to that question and is governed by the framework above regardless of how Bilinski is ultimately decided.

Identity

In a sexual assault case where the accused and complainant are known to each other, identity is rarely in issue. Where they are strangers, or where the encounter was brief, dark, or otherwise unfavourable to careful observation, identity may be the central question at trial. The Supreme Court of Canada has recognized for decades that eyewitness identification — even when given by an honest and convinced witness — is among the most frail forms of evidence in criminal law.

The reliability of identification evidence

The Supreme Court of Canada in Mezzo v. The Queen set out the matrix for assessing the reliability of identification evidence. The factors include the length and distance of observation, the lighting, any obstructions to view, whether the witness knew the accused before the encounter, the time that elapsed between the observation and the description given to the police, and discrepancies between that initial description and the accused's actual appearance. Wilson J. observed that other factors — the consistency of the witness's descriptions, her degree of attention at the time, her reaction at subsequent encounters with the accused — also bear on quality. The list is not a closed one.

The Court returned to the subject in R v Hibbert. Justice Arbour observed that the danger associated with eyewitness identification in court is that it is "deceptively credible, largely because it is honest and sincere." The dramatic effect of identification taking place in court, before the jury, can aggravate the distorted weight a jury places on it. Equally significant is what the Court called the very weak link between the confidence level of a witness and the accuracy of that witness — a confident identification is not, for that reason, an accurate one. Where the witness has seen the accused on television, in the news, at a preliminary hearing, or at a prior proceeding, her recollection of the assailant cannot easily be separated from the mental images formed on those subsequent occasions.

The factors and the warnings work together. In R v Favel, the Provincial Court of Alberta acquitted an accused charged with unlawful confinement and robbery on identity grounds. The complainant had spent twenty to thirty minutes in the accused's company and identified the accused in a photo line-up and again in court. The acquittal turned on an accumulation of frailties: the complainant's opportunity to observe the assailant's face was less complete than the time spent together suggested; the description given to the police lacked distinctive features that the accused in fact had; the complainant had earlier identified a different person from a Facebook search; and the photo line-up made the accused the only person matching the age description. The case is not exceptional. It is what the Mezzo framework produces when applied carefully to identification evidence that, on its surface, looks like it might support a conviction.

Alibi: notice and consequences of disbelief

An accused who intends to rely on the defence that he was not present at the place and time the offence is alleged to have occurred must give the Crown notice of that defence in time for the police to investigate. The Supreme Court of Canada in R v Cleghorn set out the framework. Disclosure of an alibi has two components: it must be adequate, and it must be timely. Adequate disclosure means a statement that the accused was not present at the place of the crime when it was committed, an account of the accused's whereabouts at the relevant time, and the names of any witnesses to the alibi. Timely disclosure means disclosure early enough to permit a meaningful police investigation; it need not be made at the earliest possible moment, and it need not be made by the accused personally. Third party disclosure, including disclosure by a witness to the alibi, is sufficient.

The Court was clear about what follows from a failure to give proper notice. The rule is one of expediency, not of law. Improper disclosure does not exclude the alibi evidence at trial; it permits the trier of fact to draw an adverse inference when weighing it. The reasoning is grounded in the accused's right to silence and in the recognition that an overly strict approach to disclosure would erode that right.

A different question arises where an alibi is properly disclosed, advanced at trial, and disbelieved by the trier of fact. The Supreme Court of Canada in Hibbert summarized the law on this point with precision. A disbelieved alibi, on its own, has no evidentiary value — it does not, by being disbelieved, support an inference that the accused is guilty. A disbelieved alibi is not, on its own, evidence that the alibi was fabricated. To support the inference that an alibi was deliberately fabricated, and that the accused was involved in the fabrication, there must be other evidence pointing to that conclusion. The inference of consciousness of guilt is grounded in the attempt to deceive, not in the failure of the alibi. Even where the inference is available, the jury must be told that it may, not must, be drawn. A fabricated alibi is not conclusive evidence of guilt.

Disclosure and records

A sexual offence prosecution involves more constitutional architecture around what the defence is entitled to see, and what the defence has to apply for, than almost any other kind of criminal case. The Crown's general disclosure duty applies, and it is broad. Parliament has also enacted statutory regimes that govern records of the complainant — what can be applied for, what must be applied to introduce, and what restrictions attach to materials the defence already has. Each regime has its own threshold, its own procedure, and its own constitutional history.

Disclosure as the constitutional baseline

The Crown's duty to disclose all relevant non-privileged information in its possession is the constitutional baseline. The Supreme Court of Canada in R v Stinchcombe grounded the duty in the right to make full answer and defence under section 7 of the Charter. The "fruits of the investigation" in the Crown's hands are not the Crown's property to deploy in securing a conviction; they are held in trust for the proper administration of justice, and the right to make full answer and defence is one of the "pillars of criminal justice" on which the system depends. Information must be disclosed unless it is clearly irrelevant or subject to legal privilege; the test for withholding is whether there is a reasonable possibility that doing so would impair the right to make full answer and defence.

The Supreme Court of Canada in R v McNeil closed an obvious gap. The duty under Stinchcombe rests on the Crown, but the material that produces a criminal case sits with the investigating police force. The Court held that the police force, although legally distinct from the Crown, is not a third party for disclosure purposes. It is on the same first-party footing as the Crown, with a corollary obligation to disclose to the Crown all material pertaining to its investigation of the accused. Records of serious misconduct by officers involved in the investigation come within that obligation where the misconduct is related to the investigation or could reasonably impact on the case. The Crown is the gatekeeper for what comes through to the defence, but the Court was direct about the institutional posture: the Crown is not an ordinary litigant, and the Crown and the defence are not adverse in interest on the question of whether the investigating officers in a given case are reliable.

Records of the complainant: the Mills regime

Records of the complainant — therapy notes, counselling records, medical files, personal journals, school records, child welfare records — are not handled under Stinchcombe. They are handled under a statutory regime Parliament enacted in 1997 and the Supreme Court of Canada upheld in R v Mills. The regime, in sections 278.1 through 278.91 of the Criminal Code, applies to records in respect of which the complainant has a reasonable expectation of privacy. The Court in Mills upheld it eight to one.

The architecture is not the same as the common-law procedure the Court had set out four years earlier in R v O'Connor. The common-law O'Connor procedure still applies to third-party records in non-sexual-offence proceedings. For sexual offences, the statutory regime takes over. The accused has to bring an application before the trial judge, with notice to the complainant and the record-holder. Two findings are required at the first stage before the judge will even look at the records: the records must be likely relevant to an issue at trial or to the competence of a witness, and production must be necessary in the interests of justice. The judge weighs the application against the factors in section 278.5(2), which balance the accused's right to make full answer and defence against the complainant's privacy and equality interests, society's interest in the reporting of sexual offences and the obtaining of treatment, and the integrity of the trial process. If the records pass that threshold, the judge inspects them and decides what, if anything, is produced to the accused and on what conditions.

The Court in Mills recognized that this allocation of burden puts the accused in a difficult position. The accused is being asked to demonstrate the relevance of records he has not seen. That is the "Catch-22" the Court itself named, and the same problem the Court returned to in McNeil, where it described an "impossible Catch-22 position." The majority's answer in Mills is in the discretion of the trial judge. Quoting Justice L'Heureux-Dubé's reasons in O'Connor, the Court said that "in borderline cases the judge should err on the side of production to the court." The judge's review of the records is the safeguard; the threshold is designed to keep the regime from being used for fishing expeditions, not to keep accused persons from records that could matter to their defence.

Records in the accused's possession

Where the accused already has records of the complainant — text messages, emails, photographs, social media exchanges, recordings — a different question arises. Until 2018, there was no statutory procedure governing the use of such materials at trial. Parliament added one in Bill C-51, in sections 278.92 through 278.94. The Supreme Court of Canada in R v J.J. upheld the regime six to three. An accused who proposes to introduce or use records of the complainant in which the complainant has a reasonable expectation of privacy must bring an application before trial. Where the records also contain evidence of the complainant's prior sexual activity, the admissibility analysis runs through the section 276 criteria; otherwise, it runs through the significant-probative-value test in section 278.92(2)(b). Either way, the procedure is the one set out in sections 278.93 and 278.94: a written application with detailed particulars, a screening step before the judge decides whether to hold a hearing at all, and an in-camera hearing at which the complainant is entitled to appear, make submissions, and be represented by counsel. The procedure is structurally similar to the Mills regime, but it is its own statutory scheme, directed at admissibility rather than production. The threshold for admissibility is high: the record must have significant probative value that is not substantially outweighed by the danger of prejudice to the proper administration of justice. The practical implication is that the materials a defendant has in his own possession cannot simply be put to the complainant in cross-examination or referred to in submissions without going through the process. Anyone facing a sexual offence charge who has communications or other materials concerning the complainant should expect that an application will be needed before those materials can be used at trial, and should expect to make decisions about that material with counsel well before the trial date.

What the regimes mean in practice

The four regimes — Stinchcombe, O'Connor, the section 278 records regime, and the section 278.92 admissibility regime — operate concurrently in any Canadian sexual offence prosecution. The Crown discloses what is in its hands and the hands of the investigating police force. Records of the complainant in third-party hands are accessed, if at all, through the statutory regime. Records of the complainant already in the accused's possession are introduced, if at all, through the post-2018 admissibility regime. Each has its own application, threshold, and timeline.

The doctrinal foundation is the principle the Supreme Court of Canada has repeated since well before any of these regimes were enacted: "the Crown must establish a case to meet before the accused is expected to respond." The records regimes shift some of the work of putting evidence before the trial judge onto the accused; the Court has held that they do so within constitutional limits, on the strength of the safeguards built into them. Both Mills and J.J. divided the Court — eight to one in Mills, six to three in J.J. — but the controlling rule is the majority view in each.

The practical consequence is that disclosure and records issues in a sexual offence case need pre-trial planning. Applications take time, and some have to be brought weeks or months before trial. The accused's strategic position is materially different from a non-sexual-offence prosecution, in ways that affect how the case is built well before the trial date arrives. The earlier counsel is engaged, the more useful those decisions will be.

Reverse disclosure: applications the defence must bring

The Crown's disclosure obligations under Stinchcombe, discussed in the previous section, are not the only disclosure obligations in a sexual offence prosecution. The defence has its own — and they are structurally different from the Crown's. The accused who intends to lead evidence of the complainant's prior sexual activity, or who intends to use records of the complainant already in his possession, has to apply to the trial judge before trial, on notice to the Crown and to the complainant, in a sworn application setting out the use proposed for the evidence.

Three things shape what a defendant has to apply for and when. The first is the section 276 regime governing prior sexual activity. The second is the section 278.92 admissibility regime governing records of the complainant in the accused's possession, discussed in Section 5. The third is the right to silence as it operates within these pre-trial applications under section 278.93. Each has its own threshold, its own procedure, and its own constitutional history.

Prior sexual activity: the section 276 regime

Section 276 of the Criminal Code governs the admissibility of evidence that a complainant has engaged in sexual activity other than the activity that forms the subject of the charge. The Supreme Court of Canada in R v Darrach upheld the regime against constitutional challenge in 2000. The bar in section 276(1) is categorical: evidence of prior sexual activity is inadmissible to support the inference that the complainant is more likely to have consented to the activity in question, or less worthy of belief, by reason of the sexual nature of the activity. Those are the "twin myths" identified by the Court in Seaboyer and codified in the 1992 amendments. The Court was careful about scope. The bar applies to all sexual activity, whether with the accused or with someone else, and whether or not the activity was consensual. What it does not do is exclude relevant evidence. An accused, the Court said, has never had a right to adduce irrelevant evidence; the regime controls admissibility through criteria and procedure, not through a blanket exclusion. Where evidence is offered for a use that does not invoke the twin myths, section 276(2) sets a threshold of significant probative value — read in Darrach as evidence not so trifling as to be incapable, in the context of all the evidence, of raising a reasonable doubt. The Court framed the architecture in terms it has returned to in every section 276 case since: the right to make full answer and defence does not include defending by ambush.

The Supreme Court of Canada returned to the regime in R v Goldfinch in 2019, applying it to the most common factual setting — defence-led evidence of a prior sexual relationship between the accused and the complainant. Justice Karakatsanis for the majority held that bare assertions of relevance to context, narrative, or credibility do not satisfy section 276. The accused has to positively identify a use of the proposed evidence that does not invoke twin-myth reasoning, and has to do so with precision in the written application. On the facts, the accused's "friends with benefits" framing served no purpose other than to support the prohibited inference that a complainant who had consented to sex with the accused in the past was more likely to have consented on the night in question. The Court named the categories where prior sexual activity can properly be relevant: where the accused relies on honest but mistaken belief in communicated consent and points to specific evidence of how consent was previously communicated; where the complainant has made prior inconsistent statements about the existence of the sexual relationship; and in the rare case where the absence of the evidence would render the defence narrative untenable. Generic invocations of "context" or "credibility" are not on that list.

The Supreme Court of Canada in R v Barton, in the section 276 aspect of its 2019 decision, addressed two further points. The first is who is responsible for compliance with the regime. The trial judge is the gatekeeper. The Court put it directly: "the trial judge, not the Crown, who is the gatekeeper in a criminal trial." Crown inadvertence cannot waive the dignity, privacy, and equality interests of the complainant that section 276 protects, and a trial judge faced with prior sexual activity evidence has to raise the regime even where neither party has done so. The second is that the categorical bar in section 276(1) applies regardless of which party leads the evidence. The bar binds Crown-led prior sexual activity evidence as well as defence-led evidence, and trial judges are to police Crown-led evidence under the common-law principles in Seaboyer. The decision also has a doctrinal connection to the mens rea analysis discussed in Section 2 — the Court's renaming of the mistaken belief defence to "honest but mistaken belief in communicated consent" means that prior sexual activity adduced to support that defence has to rest on evidence of how consent was previously communicated between the parties, not on the bare fact that consent had been given in the past.

The Supreme Court of Canada's most recent decision on the section 276 regime is R v Kinamore, released in 2025. The Court held that evidence of a complainant's sexual inactivity — that the complainant has not engaged in, or prefers not to engage in, certain sexual activity or sexual activity under certain circumstances — is part of sexual history and is presumptively inadmissible. The case identified what the Court called the "inverse twin myth": the inference that a complainant who has not engaged in particular sexual activity in the past was less likely to have consented, or is more worthy of belief on that account. Both inferences are impermissible. The Court also harmonized the common-law procedure for Crown-led prior sexual activity evidence with the section 276 statutory regime. A written application before trial, a two-stage hearing, publication restrictions, and a mandatory limiting instruction if the evidence is admitted at a jury trial all apply on the Crown side, with a single distinction: the admissibility standard for Crown-led evidence is the ordinary one (prejudicial effect outweighs probative value), while the standard for defence-led evidence requires that prejudicial effect substantially outweigh probative value. Where the same evidence is relied on by both parties, the defence-led standard governs.

Records of the complainant in the accused's possession

Section 5 discussed the section 278.92 admissibility regime that governs how records of the complainant already in the accused's possession — text messages, emails, photographs, social media exchanges, recordings — can be used at trial. From the standpoint of pre-trial strategy, the operative point is the timing. The application has to be brought before trial. The accused has to commit, in the written application, to a use for the material before the complainant has testified and before cross-examination has begun. The two-stage hearing produces a ruling that defines what cross-examination on the materials is possible at trial, and what is not. The decisions about what to apply for, and when, are made early — well before the trial date arrives. The Supreme Court of Canada upheld the regime in J.J. in 2022 on the strength of the safeguards built into the procedure. Those safeguards work only if the accused has counsel engaged early enough to assess the material, identify the use that has a realistic prospect of admissibility, and draft an application that meets the threshold.

The right to silence and the pre-trial application

An accused is not required to testify at trial, and the Crown bears the burden of proving every element of the offence beyond a reasonable doubt. Those propositions, discussed in Section 3, are not displaced by the section 276 and section 278.92 regimes. What the regimes do require is that the accused who wishes to lead prior sexual activity evidence, or to use records of the complainant in his possession, file a sworn application before trial setting out the use proposed for the evidence. The application is, in substance, a pre-trial commitment.

The Supreme Court of Canada in Darrach addressed the self-incrimination concern that the regime can appear to raise. The pressure on the accused to participate in the pre-trial application is, the Court held, tactical rather than legal. The accused is not legally compelled to swear an affidavit, to testify at the voir dire, or to identify in advance the use he proposes to make of the evidence. He is, however, tactically compelled to do so if he wants the evidence admitted. The Court framed that distinction directly and held that tactical pressure of this kind does not engage the principle against self-incrimination or the right to a fair trial. The architecture of the regime is what it is because the alternative — admitting prior sexual activity evidence or records of the complainant without pre-trial screening — would, on the Court's reasoning in Darrach and J.J., fail to protect the privacy, dignity, and equality interests the regimes were enacted to safeguard.

The practical consequence is that decisions about what to apply for, when, and on what basis are made early in a sexual offence file and have downstream implications for how the trial unfolds. An accused choosing whether to bring an application, and on what evidence, is making decisions that shape what defence is available at trial. Those decisions, like the decisions discussed in Section 3 about whether and how to testify, are best made with counsel who has read the disclosure, knows the file, and has spoken with the accused at length.

What to expect if convicted

Sexual assault under s. 271 of the Criminal Code is a hybrid offence. On indictment, the maximum is ten years in prison; where the complainant is under sixteen, the maximum is fourteen years and a one-year minimum applies. The more serious offences — sexual assault with a weapon or causing bodily harm under s. 272, and aggravated sexual assault under s. 273 — carry maximums of fourteen years and life respectively, with minimums in certain circumstances. Election, mode of trial, and the Crown's view of which offence captures the conduct alleged all bear on what a conviction would actually look like in custody terms.

Within those statutory ranges, sentencing in Alberta is shaped by appellate guidance. For a "major sexual assault" — non-consensual vaginal intercourse, and other acts the Court of Appeal has treated as equally serious — the Alberta Court of Appeal in R v Arcand reaffirmed three years' imprisonment as the starting point. Lesser sexual assaults sit in different ranges. The Supreme Court of Canada in R v Lacasse and R v Parranto has more recently emphasized that starting points are guidance rather than presumptive sentences, and that the trial judge's task remains an individualized application of the sentencing principles in ss. 718, 718.1, and 718.2 of the Criminal Code. What actually happens at sentencing on a given file depends on the conduct proved, the offender's circumstances, any aggravating or mitigating factors, and the trial judge's view of where the case sits against the appellate guidance.

A conviction for a sexual offence triggers a set of ancillary orders that are mostly automatic on the Crown's application, regardless of the sentence imposed. Registration under the Sex Offender Information Registration Act (SOIRA) follows on the framework Parliament enacted in 2023 in Bill S-12, with the duration of registration set by reference to the offence and the sentence under ss. 490.012(1), (2), and (3). A DNA order under s. 487.051 is mandatory on conviction for any sexual offence designated as a primary offence under s. 487.04. A firearms prohibition order under s. 109 or s. 110 is also typically engaged. Where the complainant was under sixteen, a s. 161 prohibition order — restricting contact with minors, presence at certain places, and access to certain employment — may also apply. Each of these orders has its own duration, scope, and review mechanism; the right answer on a given file requires case-specific advice.

A conviction also carries consequences that sit outside the sentence itself. A criminal record affects employment, professional licensing, volunteer work involving vulnerable people, travel to the United States and elsewhere, and — for anyone who is not a Canadian citizen — immigration status under the Immigration and Refugee Protection Act. Immigration consequences of a sexual offence conviction can be severe and are not addressed by the criminal court at sentencing; they require their own assessment. The Supreme Court of Canada has recognized that collateral consequences of this kind are relevant at sentencing, but they do not displace the sentencing range — they inform where, within that range, the sentence properly sits. Anyone facing a sexual offence charge should have a frank early conversation with counsel about what a conviction would actually mean across the categories above before making decisions about plea, election, or trial strategy.

We act on sexual offence files in the Alberta Court of Justice, the Court of King's Bench of Alberta, and the Court of Appeal of Alberta. From our Calgary office we travel to court in Airdrie, Cochrane, Canmore, Okotoks, Diamond Valley, Strathmore, Didsbury, Drumheller, Hanna, Brooks, Medicine Hat, Lethbridge, Fort Macleod, Taber, Pincher Creek, and Cardston, and we appear in matters arising on the Tsuut'ina, Siksika, and Kainai (Blood Tribe) Nations.

KJR Law is the trade name under which Brad Kraus, Greg Janzen, and Elena Ryland practise in association as independent practitioners. They are not partners. Each lawyer is independently responsible for their own clients.

Initial consultation at no charge. Scope is limited to an initial assessment of your matter and is subject to a conflict check. GST does not apply to a no-charge consultation.

Contact KJR Law

The decisions described on this page — whether to apply under section 276 or section 278.92, whether and how to testify, what to do with records the defence has in its possession, how to assess identification evidence, what a guilty plea or a contested trial would actually mean — are not decisions to make alone, and they are not decisions to make late. Many of them have to be made well before trial. Some have to be made within days of charges being laid. Bail conditions, decisions about how the charge will be prosecuted, the review of disclosure as it arrives, and the early assessment of which defences are realistic on the facts all sit in the period between charge and trial — and the period is often shorter than people expect. An accused who has read this page and is trying to assess where his own case sits should speak with criminal defence counsel sooner rather than later.

KJR Law is a Calgary criminal defence firm. Sexual offence prosecutions are part of the practice. A first conversation can clarify what stage the matter is at, what decisions are approaching, and what options are realistic on the facts as they stand. The firm can be reached by phone at 587-583-2808, or through the contact page at kjr-law.ca/contact.