The criminal justice system can be daunting, but you don’t need to go through it alone. Our Criminal lawyers are here to guide you every step of the way.
Contact Our Firm
Case Commentary · Criminal Law · Evidence
Ontario’s Court of Appeal overturns two sexual assault convictions, finding that a trial judge improperly used an accused’s marital infidelity to infer a propensity to disregard a complainant’s wishes.
R. v. S.F., 2026 ONCA 293 · April 23, 2026 · van Rensburg, Gomery & Pomerance JJ.A.
R. v. S.F., 2026 ONCA 293 · Court of Appeal for Ontario · COA-24-CR-0607
On appeal from convictions entered by Justice Scott G. Pratt, OCJ, February 1, 2024.
By Joseph Neuberger, Michael Bury and Diana Davison, Neuberger & Partners, Toronto Criminal Lawyers
The appellant, S.F., was convicted at trial of three counts of sexual assault arising from two incidents with his former romantic partner. The complainant testified that on two separate occasions, the appellant penetrated her vaginally without a condom, contrary to her clearly communicated condition for consent. The appellant denied the conduct entirely, maintaining that the complainant had consented to all sexual activity.
The trial judge disbelieved the appellant, citing evasive and inconsistent testimony. The appellant received a suspended sentence on the first count, and 18 months’ imprisonment consecutive on each of the second and third counts — a total of three years in jail. He appealed the convictions on counts two and three.
The trial judge’s pivotal error lay in how he used evidence that the appellant had been unfaithful during his relationship with the complainant. The infidelity first surfaced in the complainant’s own testimony as context for their breakup. But in his reasons, the judge went further — he used it to draw an inference about the appellant’s character:
“This shows a lack of respect for both the Complainant and their relationship. It shows he did not take their commitment or her feelings seriously. That is relevant to my determination of whether he would disregard the Complainant’s rule regarding condom use.”
At first glance, this reasoning might appear contextual rather than propensity based. The judge was careful to say he was not using infidelity as general bad character evidence. But the Court of Appeal was not persuaded by that caveat. As the court found at paragraph 6, the reasoning “could only be seen” as propensity reasoning — the judge was using the appellant’s disrespect in one context (monogamy) to infer a predisposition to disrespect in another (condom use).
That is precisely the chain of reasoning that Canadian evidence law forbids. Under R. v. Handy, evidence of prior discreditable conduct cannot be used to show that an accused is the kind of person who commits bad acts, and therefore likely committed the offence charged. People are tried for what they did, not who they are.
The court identified a second, compounding problem: the appellant had no notice that his admitted infidelity could be turned against him as propensity evidence. He was not cross-examined on whether his unfaithfulness reflected a general disregard for the complainant’s boundaries. The Crown did not advance the argument in closing. The trial judge did not invite submissions on the point before deploying the reasoning.
This is a significant issue because accused persons are entitled to understand the case they have to meet. An argument that a defendant’s moral failing in one area of life makes them more dangerous in another is not a trivial inference — it is a serious one that requires scrutiny and response. The appellant was denied that opportunity entirely.
The Crown attempted to salvage the convictions by invoking R. v. Kruk, 2024 SCC 7, arguing that a trial judge may legitimately find an accused “generally disinclined” to respect a complainant’s wishes based on prior conduct. The Court of Appeal rejected this reading firmly at paragraph 9.
In Kruk, the Supreme Court upheld a judge’s rejection of an accused’s account of a specific evening because it was inconsistent with the judge’s other findings about events on that same evening. That is orthodox credibility assessment — testing an account for internal consistency. It does not authorize reasoning that a morally reprehensible act committed in the past makes a person more likely to commit an unrelated criminal offence months later. The Crown had mischaracterized the passage it relied upon.
When an error is found at trial, an appellate court can nonetheless dismiss the appeal if it concludes that no substantial wrong or miscarriage of justice resulted — the so-called curative proviso under s. 686(1)(b)(iii) of the Criminal Code. The Crown argued the proviso should apply because the trial judge had other reasons for disbelieving the appellant beyond the infidelity.
The Court of Appeal declined. The problem is not merely that a bad reason was mixed in with good ones. The W.(D.) framework — the three-step analysis governing how trial judges assess accused testimony — requires a careful, structured consideration of credibility. The court had no way of knowing how much the prohibited propensity reasoning infected the judge’s overall findings at each step of that framework. Moreover, the procedural unfairness caused by the lack of notice was itself an independent basis for intervention.
Convictions on counts 2 and 3 set aside. New trial ordered on those charges. Sentence appeal dismissed as abandoned. Three years’ imprisonment cannot stand.
R. v. S.F. is a clear application of settled principles, but its value lies in the warning it sends to trial judges handling cases where parties had complicated relationships. Evidence of bad behaviour in a relationship — dishonesty, infidelity, controlling conduct — often enters the record for legitimate purposes, such as explaining the context of a breakup or addressing a motive to fabricate. Judges must be vigilant that such evidence does not quietly migrate into the credibility analysis as a proxy for bad character.
The case also reinforces the procedural dimension of evidence law. Even if a creative argument for using evidence in a novel way exists, fairness requires that the accused be given the chance to respond to it. Surprise propensity reasoning — appearing for the first time in the reasons for judgment — cannot be cured by pointing to other findings in the verdict.
The complainant’s allegations will now be assessed at a fresh trial, before a new trier of fact, on a permissible evidentiary basis.
If you are facing a serious sexual or domestic assault related offence, particularly arising from the end of a relationship, contact Neuberger & Partners LLP, Criminal Lawyers Toronto, an award-winning firm specializing in these cases for over 30 years.
Review our Newsletter on Sexual Assault Case Law and our podcast – Not on Record – to learn more.
This post is a legal commentary for informational purposes only. It is not legal advice. This appeal is subject to a publication ban under s. 486.4 of the Criminal Code; the parties are identified only by initials in the decision.
