Supreme Court tightens use of criminal records but upholds murder conviction

The Supreme Court of Canada has clarified when an accused person’s criminal record can be used to challenge their testimony, warning that a history of offending does not, by itself, establish a specific willingness to lie. The ruling leaves Awale Hussein’s second-degree murder conviction intact.

In R. v. Hussein, 2026 SCC 2, released January 23, all nine judges dismissed the Ontario appeal. Chief Justice Richard Wagner, writing for seven judges, found serious errors in the decision to admit Hussein’s entire record. The majority nevertheless upheld the conviction because the other evidence was overwhelming. Justice Mahmud Jamal, joined by Justice Malcolm Rowe, reached the same result but found no reviewable error in the trial ruling.

A criminal record is not proof of guilt

The case arose from the fatal stabbing of Brian Boucher at an Ottawa apartment after a night of drinking on February 1, 2017. No one witnessed the stabbing. The trial concerned both the identity of the killer and the mental state required for murder. Hussein sought to prevent or limit questioning about his previous offences, but the trial judge dismissed that application and the Ontario Court of Appeal upheld his conviction.

The application was governed by the framework established in R. v. Corbett in 1988. Section 12(1) of the Canada Evidence Act permits questioning about prior convictions to assess a witness’s credibility, including when an accused testifies. Judges retain discretion to exclude convictions when their prejudice to a fair trial outweighs their value in assessing whether the accused’s testimony can be trusted.

The majority rejected Hussein’s request to make those convictions presumptively inadmissible. Parliament’s permissive wording remained in force, and Hussein had not brought a new constitutional challenge to invalidate it. An accused seeking exclusion must persuade the judge, on a balance of probabilities, that the prejudice outweighs the evidence’s value for credibility.

That starting point does not make admission automatic. The majority stressed the danger that jurors will treat a criminal record as evidence that the accused is the kind of person likely to commit the charged offence. Convictions admitted for credibility cannot be used that way. Judges should exclude records lacking sufficient evidentiary value rather than assume that an instruction to the jury will resolve the problem.

Honesty, social context and youth records

The majority said the nature of a previous offence is the most important factor in assessing its value for credibility. To carry enough weight to overcome its prejudice, a conviction must support an inference of a specific capacity or willingness to be dishonest. Perjury, fraud and forgery are examples; some offences involving breaches of court promises or undertakings can also bear on whether a witness will respect an oath.

Many violent offences reveal little about honesty. The majority rejected using general disregard for the law, or the sheer length of a record, as a substitute for the required assessment. It nevertheless stopped short of a categorical ban on violent convictions: robbery, for example, combines violence and dishonesty, and other features of a trial may affect admissibility.

Social context also matters. Where a prior conviction stems partly from poverty, discrimination or other disadvantage, its value as an indicator of dishonesty may be reduced. The accused need not establish direct causation, but must show some connection between that disadvantage and the circumstances leading to the conviction. Judges can use testimony and judicial notice of established social conditions without necessarily requiring professional reports.

The majority also recognized that racist stereotypes and other discriminatory biases can increase the prejudice caused by admitting a record. Social context does not determine the outcome by itself; the assessment must address the individual accused and the circumstances in which the record accumulated.

Youth offences generally carry less weight in assessing an adult’s credibility because young people’s maturity, independence and judgment are still developing. Older convictions also tend to lose probative force over time. Conversely, strong similarity between a previous offence and the charge being tried increases the risk of prejudice and calls for particular caution.

The court also narrowed when concerns about a distorted comparison between witnesses become relevant. The defence must have attacked an important Crown witness’s credibility through alleged bad character unrelated to the case’s facts. Merely exposing inconsistencies or reasons to disbelieve a witness on the evidence does not create a corresponding entitlement to introduce the accused’s record.

Even when such a character attack creates a genuine risk of distortion, it does not open the door to the entire record. Judges must weigh the scale of the attack, the importance of the witness and other evidence already before the jury, and consider whether admitting only part of the record would address the problem.

Serious errors, but no new trial

Applying those principles, the majority concluded that Hussein’s youth offences and his prior offences of assault, uttering threats and possession of a weapon should have been excluded. The trial judge had overstated their value for credibility, understated their prejudice and placed too much weight on the defence’s attack on a Crown witness.

The majority also held that the strength of the Crown’s case is irrelevant when deciding the Corbett application. Strong evidence does not make it permissible for jurors to draw an improper inference from past offending. That question is distinct from the later appellate inquiry into whether a conviction can survive an established legal error.

Under the Criminal Code’s curative proviso, the Crown must show why a conviction should stand despite an error. One route concerns minor or harmless mistakes; another concerns serious errors where the evidence is so overwhelming that conviction was inevitable. The majority expressly rejected treating the errors in Hussein’s trial as harmless, but found that the second route applied.

The majority pointed to forensic evidence connecting Hussein to the killing, including his blood at several locations in the apartment, a healing thumb wound and DNA evidence on a toque found close to where Boucher died. It also relied on his departure from the scene and subsequent conduct. The repeated knife attack and evidence of purposeful actions afterward supported the conclusion about his mental state, despite evidence of intoxication.

The threshold for applying the proviso on that basis is exceptionally high. The majority described this as one of the rare cases meeting it and concluded that there was no reasonable possibility of a different verdict without the trial judge’s errors. Hussein therefore received no new trial.

Two judges would leave the framework unchanged

Jamal and Rowe agreed that the conviction should stand, but disagreed with the majority’s changes to the Corbett analysis. In their view, a pattern of contempt for the law and the strength of the Crown’s case could remain relevant considerations. They emphasized trial judges’ discretion and the ability of juries to follow instructions limiting how a record may be used.

The concurring judges did not endorse automatic admission of lengthy records. They called for a contextual assessment and care in using the strength of the prosecution’s case. They found no reviewable error here and would have deferred to the trial judge. The majority’s more restrictive guidance governs future Corbett assessments, while the unanimous dismissal preserves Hussein’s conviction.

Source: R. v. Hussein, 2026 SCC 2.