Citation: R. v. Berg, 2026 SCC 21 Decision Date: June 5, 2026 Court: Supreme Court of Canada (9-0)
Background
Matthew Berg was convicted of sexual assault following a trial in Saskatchewan. The central factual dispute was consent. The complainant testified she did not consent and Berg testified the encounter was entirely consensual. The trial judge had no identified frailties in Berg’s evidence when considered in isolation, but rejected it anyway, stating: “I reject Mr. Berg’s account because I accept [the complainant’s] testimony.”
Berg argued on appeal that the trial judge had committed two errors: failing to resolve material inconsistencies in the complainant’s evidence, and improperly using her evidence to reject his account without assessing it against the totality of the evidence. The Saskatchewan Court of Appeal dismissed the appeal by majority. The SCC unanimously agreed the conviction should stand, but wrote to correct a widespread analytical error it had observed in trial courts across the country.
The W.(D.) Framework: A Brief Refresher
R. v. W.(D.), [1991] 1 SCR 742 remains the foundational framework for credibility assessment in criminal trials. It prevents trials from collapsing into simple credibility contests. The core principle: a conviction cannot rest solely on the trier of fact preferring or believing the Crown’s witness. If the accused’s evidence, or the evidence as a whole, raises a reasonable doubt, the accused must be acquitted. Belief in the complainant is not equivalent to proof beyond a reasonable doubt.
The Problem: J.J.R.D. Misapplied at Trial
In R. v. J.J.R.D. (2006), 218 OAC 37, Doherty J.A. of the Ontario Court of Appeal wrote a passage that has since taken on a life of its own. Addressing a sufficiency-of-reasons argument on appeal, he observed that a trial judge can reject an accused’s evidence based on “a considered and reasoned acceptance beyond a reasonable doubt of the truth of conflicting credible evidence” meaning appellate courts could confirm a conviction where the trial judge had adequately explained, through reference to compelling and credible Crown evidence, why the accused’s account was rejected.
The problem: trial judges began invoking this passage as a template for structuring their own W.(D.) analyses. It was used to justify rejecting facially unassailable defence evidence simply by articulating a “considered and reasoned” acceptance of the complainant’s testimony, effectively treating credibility preference as a substitute for proof beyond a reasonable doubt.
The SCC in Berg identified this as a serious and recurring error, noting a string of ONCA decisions grappling with exactly this misapplication.
What the Court Held
The Court’s guidance in Berg is clear and unambiguous:
1. J.J.R.D. is an appellate sufficiency-of-reasons case… nothing more.
It was designed to help appellate courts assess whether a trial judge had adequately explained a conviction, not to provide trial judges with a formula for convicting. The passage from Doherty J.A. describes what kind of explanation suffices for appellate review; it does not describe what kind of reasoning suffices to ground a conviction.
2. Trial judges should not rely on J.J.R.D. when structuring W.(D.) analyses or giving W.(D.) jury directions.
The Court was categorical on this point. J.J.R.D. does not offer a formula for overcoming facially unassailable exculpatory evidence. Invoking it in a trial setting risks misleading the trier of fact into thinking that a “considered and reasoned” acceptance of the complainant’s evidence is sufficient to reject the accused’s account, which it is not.
3. Conviction requires proof beyond a reasonable doubt and not credibility preference.
The Court reaffirmed that W.(D.) exists precisely to prevent trials from becoming credibility contests. The presumption of innocence means a guilty verdict cannot rest on whether the trier of fact believes the Crown’s evidence or finds it more plausible than the defence’s. Proof beyond a reasonable doubt demands more than a considered acceptance of the complainant’s testimony.
4. The full J.J.R.D. passage, properly read, already requires the BARD standard.
Doherty J.A.’s formulation in fact included the words “beyond a reasonable doubt” and “a considered and reasoned acceptance beyond a reasonable doubt of the truth of conflicting credible evidence.” The Court in Berg emphasized this qualifier. When that passage is read in full and in context, it reinforces rather than undermines the BARD requirement. Trial judges who stripped out those words and applied the truncated version were misreading the case from the start.
Why Berg’s Conviction Was Upheld
Despite the legal guidance, the appeal was dismissed because the trial judge in Berg’s case did not actually engage in a naked credibility contest. A functional reading of his reasons revealed that he assessed Berg’s account against the complainant’s credible testimony and corroborating circumstantial evidence, and only then was satisfied of guilt beyond a reasonable doubt. The problematic statement: “I reject Mr. Berg’s account because I accept [the complainant’s] testimony” was imprecise language, but not a true reflection of the analytical process the reasons as a whole disclosed.
This is an important limiting principle: appellate courts review reasons functionally, not formulaically. A poorly worded passage does not automatically vitiate an otherwise sound analysis.
Practical Implications for Defence Counsel
Berg is immediately useful on appeal and, arguably, at trial.
On appeal, where a trial judge has structured W.(D.) by invoking J.J.R.D. as a formula, treating the complainant’s credible testimony as sufficient in itself to reject the accused’s account, that reasoning is now clearly identified as legally erroneous by the SCC. Where the error is material, it should ground appellate intervention.
At trial, Berg reinforces the W.(D.) argument in its most robust form. Defence counsel should resist any framing that reduces the trial to a question of who is believed. The trial judge must grapple with whether the accused’s evidence, or the evidence as a whole, raises a reasonable doubt, independent of whether the complainant is found credible or compelling.
In jury trials, counsel should be alert to W.(D.) instructions that echo J.J.R.D. language. The SCC has now explicitly held that J.J.R.D. should not inform jury directions. An instruction that suggests jurors may convict upon a “considered and reasoned” acceptance of the complainant’s testimony would be a misdirection.
Key Takeaway
Berg does not change the law, it corrects a widespread misapplication of it. W.(D.) remains the governing framework. What the SCC has done is close a gap that had been quietly expanding: the use of J.J.R.D. to launder credibility preference into proof beyond a reasonable doubt. For defence lawyers in sexual assault and any other credibility-driven trial, Berg is a welcome and important clarification.
William Jaksa is a criminal defence lawyer practising in Toronto, Ontario.
