Law360 Canada (October 27, 2026, 11:17 AM EDT) --
 |
| John L. Hill |
The Ontario Court of Appeal has ordered a new trial for a man convicted of two sexual assaults, an assault with a weapon and an assault against his former spouse, finding that the trial judge’s assessment of the evidence was fundamentally flawed. The decision,
R. v. Hornsby, 2026 ONCA 604, reinforces an important principle in criminal trials: a judge cannot apply different standards of scrutiny to a complainant’s evidence and an accused’s evidence, particularly when credibility is central to determining guilt.
Everton Hornsby received a global sentence of 42 months’ imprisonment on March 15, 2022. He challenged all four convictions. The offences allegedly occurred between December 2018 and June 2019, during the breakdown of the couple’s marriage.
In allowing the appeal, the court concluded that the trial judge had subjected the defence evidence to materially greater scrutiny than the Crown’s evidence. Her approach resulted in a flawed assessment of the credibility and reliability of both principal witnesses, rendering the trial unfair. The court quashed all four convictions and ordered a new trial.
Alena Butusava: ISTOCKPHOTO.COM
The couple met in Toronto in 2014 and married in 2017. The complainant sponsored Hornsby and his two children to immigrate to Canada. Their relationship deteriorated after the complainant admitted to infidelity in late 2018. They agreed their marriage was over in May 2019 but continued living in the same house.
On June 29, 2019, the complainant called police after Hornsby broke down a bedroom door. He testified that he was trying to retrieve his passport. Police arrested him for mischief. Two days later, the complainant reported the incidents that eventually led to the four convictions.
The couple’s accounts differed substantially. The trial therefore depended heavily on the judge’s assessment of their credibility and the reliability of their testimony.
The appeal did not turn on whether a trial judge may generally assess the credibility of witnesses. Rather, it concerned how those assessments were made and whether the judge applied a consistent analytical standard to the competing accounts.
The court relied on
R. v. G.F., 2021 SCC 20 and the Ontario Court of Appeal’s decisions in
R. v. Kiss, 2018 ONCA 184 and
R. v. Howe (2005), 192 C.C.C. (3d) 480.
Uneven scrutiny occurs when a trial judge assesses the credibility or reliability of defence evidence using a different degree of scrutiny than that applied to Crown evidence, resulting in an unfair trial. It provides an overarching framework for considering related errors, including inadequate reasons, misapprehension of evidence, reversal of the burden of proof and unreasonable verdicts.
The threshold is high. As the Supreme Court recognized in
G.F., credibility findings belong primarily to the trial judge, who has the advantage of seeing and hearing the witnesses. An appeal cannot simply become an opportunity to reassess credibility because the accused disagrees with the verdict.
Under
Howe, the appellant must identify something in the reasons or the record demonstrating that different standards were applied. The question is whether the judge employed a demonstrably flawed method of assessing the evidence that affected the credibility determination.
The court found that the threshold had been met.
One sexual assault conviction arose from an incident on the couple’s porch in May or June 2019. Both witnesses agreed that Hornsby ejaculated on the porch, but they disputed how it happened.
The complainant testified that Hornsby returned home during the night and ejaculated on her without consent. Hornsby denied ejaculating on her. He said she had taunted him to masturbate and that he did so some distance away while she smoked marijuana on the porch.
The trial judge accepted the complainant’s account and rejected Hornsby’s explanation as concocted. The Appeal Court found that her strong disapproval of Hornsby’s conduct had influenced her assessment of his evidence. Describing the episode as dehumanizing and absurd, she effectively allowed her condemnation of the act to foreclose meaningful consideration of Hornsby’s explanation.
The judge also treated his references to the complainant’s marijuana use as attempts to besmirch her character, although the complainant acknowledged using marijuana on other occasions. The evidence was potentially relevant to Hornsby’s account of why she was on the porch at that hour.
Further, the judge gave Hornsby no credit for admitting conduct that was against his interests, even though she treated the complainant’s admissions against interest as enhancing her credibility. These were not merely differences of opinion about the weight of evidence. They illustrated the unequal standards used to assess the two witnesses.
The court identified similar problems in the remaining counts.
The assault with a weapon conviction arose from a dispute involving a laundry basket. The complainant said she threw the basket at Hornsby and that he struck her arm with it. He testified that she hit him in the face with the basket, bloodying his nose, and that he used it defensively when she tried to strike him with a mirror.
The complainant produced a photograph of a scraped arm and elbow. However, she had not mentioned to police that her finger was broken; that allegation emerged only at trial. No photograph or medical documentation supported the alleged fracture. The trial judge failed to consider the significance of this late-emerging evidence and appeared to accept the complainant’s account principally because of the photograph.
The second sexual assault conviction concerned intercourse in the complainant’s bedroom in late June 2019. The complainant initially testified that Hornsby had been sleeping in the basement because she kept her bedroom door locked. Under cross-examination, she acknowledged that he might already have been sleeping and cuddling with her that evening, although she maintained that she had not consented to intercourse.
The Appeal Court found that the trial judge failed to analyze this inconsistency or explain how it affected the complainant’s credibility. Instead, she relied on her general view that the complainant did not want sexual relations during the separation and imported her findings about the porch incident into her assessment of the later encounter. She also failed to meaningfully address the appellant’s evidence that the intercourse was consensual.
The final conviction arose from an altercation in December 2018. The complainant testified that Hornsby slapped her thigh after becoming upset about her infidelity, leaving a handprint that she photographed. The appellant said she slapped him and kicked him first, and that he slapped her leg to stop her from kicking him. Again, the trial judge relied on the photograph and dismissed the appellant’s evidence without meaningful analysis.
The court emphasized that each charge involved a distinct incident. Although a judge may make general assessments of a witness’s credibility, those assessments cannot replace the analysis required for each count. The Crown must prove every offence beyond a reasonable doubt.
Here, the trial judge’s broadly positive assessment of the complainant and negative assessment of the accused became a fixed lens through which she evaluated the individual allegations. She treated the complainant’s admissions as signs of candour while dismissing comparable evidence from Hornsby, magnifying weaknesses in his testimony and failing to grapple with material difficulties in hers.
The decision is significant because it demonstrates that deference to trial-level credibility findings is not absolute. Appellate courts will intervene where the reasoning process itself is flawed, rather than merely because another interpretation of the evidence is possible. It also illustrates why photographs or other corroborating evidence cannot automatically resolve a credibility dispute. A photograph may establish that an injury existed, but it does not necessarily establish how the injury occurred or who was responsible. The surrounding evidence and competing explanations must still be assessed.
Nor does the decision establish that a complainant’s evidence must be treated with suspicion, or that inconsistencies necessarily undermine credibility. The requirement is even-handed analysis: material inconsistencies must be considered, explanations must be evaluated, and evidence supporting either account must be assessed without resort to stereotypes or assumptions.
The new trial order does not determine that Hornsby is innocent or that the complainant’s allegations are untrue. It means that the convictions cannot stand because the evidence was not assessed through a fair and consistent process.
The central lesson is straightforward. A criminal trial must determine whether the Crown has proved each charge beyond a reasonable doubt, not whether the judge finds one witness generally more appealing or persuasive than the other. Where unequal scrutiny distorts that determination, a miscarriage of justice has occurred and a new trial is required.
John L. Hill practised and taught prison law until his retirement. He holds a JD from Queen’s and an LLM in constitutional law from Osgoode Hall. He is also the author of Pine Box Parole: Terry Fitzsimmons and the Quest to End Solitary Confinement (Durvile & UpRoute Books). His most recent book, Acts of Darkness
(Durvile & UpRoute), was shortlisted as one of five nominees for the Crime Writers of Canada’s Brass Knuckles Award for Best Nonfiction Crime Book.
Contact him at johnlornehill@hotmail.com.
The opinions expressed are those of the author(s) and do not necessarily reflect the views of the author’s firm, its clients, Law360 Canada, LexisNexis Canada or any of its or their respective affiliates. This article is for general information purposes and is not intended to be and should not be taken as legal advice.
Interested in writing for us? To learn more about how you can add your voice to Law360 Canada, contact Analysis Editor Peter Carter at peter.carter@lexisnexis.ca or call 647-776-6740.