Credibility. R v R.G.
In (Ont CA, 2021) the Court of Appeal considered the standard of review for findings of credibility (here in a criminal case):
Principles governing appeals challenging a trial judge’s assessment of credibility. Focal Elements Ltd. v Alvand
 A trial judge’s findings of credibility are owed deference. As recently noted by the SCC in R. v. G.F., 2021 SCC 2020, 459 D.L.R. (4th) 375 at para. 81:
…a trial judge’s findings of credibility deserve particular deference. While the law requires some articulation of the reasons for those findings, it also recognizes that in our system of justice the trial judge is the fact finder and has the benefit of the intangible impact of conducting the trial. Sometimes, credibility findings are made simpler by, for example, objective, independent evidence. Corroborative evidence can support the finding of a lack of voluntary consent, but it is of course not required, nor always available. Frequently, particularly in a sexual assault case where the crime is often committed in private, there is little additional evidence, and articulating reasons for findings of credibility can be more challenging. Mindful of the presumption of innocence and the Crown’s burden to prove guilt beyond a reasonable doubt, a trial judge strives to explain why a complainant is found to be credible, or why the accused is found not to be credible, or why the evidence does not raise a reasonable doubt. But, as this Court stated in Gagnon, at para. 20:
Assessing credibility is not a science. It is very difficult for a trial judge to articulate with precision the complex intermingling of impressions that emerge after watching and listening to witnesses and attempting to reconcile the various versions of events.
In Focal Elements Ltd. v Alvand (Div Ct, 2021) the Divisional Court commented on the deference accorded credibility findings in appeals:
 Of all the factual issues determined by a trier of fact, findings of credibility attract the greatest deference. Findings of fact grounded in credibility assessments are, therefore, particularly difficult to disturb on appeal. Credibility assessments are inherently partly subjective and reflect the life experience of judges and their own perceptions. Credibility assessments are also grounded in numerous, often unstated, considerations which only the trial judge can appreciate and calibrate: Waxman v. Waxman, 2004 CanLII 39040 (ON CA) at para. 359.. Berger v. Legal Aid Ontario
In Berger v. Legal Aid Ontario (Div Ct, 2021) the Divisional Court commented on credibility and the related need for an oral hearing:
 Where a decision turns on credibility, a decision maker should not make an adverse finding of credibility without affording the affected person an oral hearing: Khan v. University of Ottawa (1977), 1997 CanLII 941 (ON CA), 34 O.R. (3d) 535,  O.J. No. 2650 (C.A.), at paras. 21-22. . R. v. R.C.
In R. v. R.C. (Ont CA, 2021) the Court of Appeal considered the appellate standard applicable to a judge's assessment of the credibility of witnesses:
 In R. v. R.A., 2017 ONCA 714, 421 D.L.R. (4th) 100, this court reviewed the principles governing an appeal based on the trial judge’s assessment of credibility and resolution of testimonial inconsistencies. This court held, at para. 46, that “an appellate court should not interfere with a trial judge’s findings of credibility if the core of the complainant’s allegations against an appellant remain largely intact on a review of the entirety of the evidence”.. R. v. G.F
 Moreover, in R. v. A.A., 2015 ONCA 558, 337 O.A.C. 20, at para. 124, this court observed that:
Disagreement with the weight assigned does not ascend to the level of palpable and overriding error. Nor does the failure to consider inconsistencies on the periphery compel intervention in the absence of any obligation to consider and respond to each claim no matter how distant from the core issues of this case. In R. v. G.F., at paras. 81-82, the Supreme Court emphasized the deference owed to a trial judge’s credibility findings, the presumption that trial judges must be taken to know the law especially “settled principles,” and that reasons which are “imperfect” or language which is “ambiguous” does not per se require reversal.
In R. v. G.F (SCC, 2021) the Supreme Court of Canada contrasts credibility and reliability:
 In R. v. Kishayinew, 2020 SCC 34, rev’g 2019 SKCA 127, 382 C.C.C. (3d) 560, and R. v. Slatter, 2020 SCC 36, rev’g 2019 ONCA 807, 148 O.R. (3d) 81, this Court adopted the reasons of a dissenting judge holding that the trial judge did not err in conflating credibility and reliability. In both cases, the trial judges accepted the complainants’ evidence and found them to be credible, even if their reliability findings were not explicit on the face of the reasons.. R. v. Borel
 As Slatter demonstrates, a trial judge’s findings of credibility deserve particular deference. While the law requires some articulation of the reasons for those findings, it also recognizes that in our system of justice the trial judge is the fact finder and has the benefit of the intangible impact of conducting the trial. Sometimes, credibility findings are made simpler by, for example, objective, independent evidence. Corroborative evidence can support the finding of a lack of voluntary consent, but it is of course not required, nor always available. Frequently, particularly in a sexual assault case where the crime is often committed in private, there is little additional evidence, and articulating reasons for findings of credibility can be more challenging. Mindful of the presumption of innocence and the Crown’s burden to prove guilt beyond a reasonable doubt, a trial judge strives to explain why a complainant is found to be credible, or why the accused is found not to be credible, or why the evidence does not raise a reasonable doubt. But, as this Court stated in Gagnon, at para. 20:
Assessing credibility is not a science. It is very difficult for a trial judge to articulate with precision the complex intermingling of impressions that emerge after watching and listening to witnesses and attempting to reconcile the various versions of events. Credibility findings must also be assessed in light of the presumption of the correct application of the law, particularly regarding the relationship between reliability and credibility. The jurisprudence often stresses the distinction between reliability and credibility, equating reliability with the witness’ ability to observe, recall, and recount events accurately, and referring to credibility as the witness’ sincerity or honesty: see, e.g., R. v. H.C., 2009 ONCA 56, 244 O.A.C. 288, at para. 41. However, under a functional and contextual reading of trial reasons, appellate courts should consider not whether the trial judge specifically used the words “credibility” and “reliability” but whether the trial judge turned their mind to the relevant factors that go to the believability of the evidence in the factual context of the case, including truthfulness and accuracy concerns. A trial judge’s determination to accept or believe inculpatory witness evidence includes an implicit assessment of truthfulness or sincerity and accuracy or reliability: Vuradin, at para. 16. Often, the term “credibility” is used in this broader sense to mean the believability of the evidence and it necessarily includes both truthfulness and accuracy: McWilliams’ Canadian Criminal Evidence (5th ed. (loose-leaf)), vol. 3, at pp. 30‑1 and 30‑2. For example, Black’s Law Dictionary (11th ed. 2019), at p. 463, defines credibility as “[t]he quality that makes something (as a witness or some evidence) worthy of belief” and model jury instructions include both truthfulness and accuracy within “credibility” assessments: G. A. Ferguson and M. R. Dambrot, CRIMJI: Canadian Criminal Jury Instructions (4th ed. (loose‑leaf)). Provided trial judges turn their mind to these considerations, there is no requirement that they utter the word “reliable”.
In R. v. Borel (Ont CA, 2021) the Court of Appeal held against the propriety of allowing credibility opinions of one witness by another:
 It is a fundamental principle of our trial process that the ultimate conclusion as to the credibility or truthfulness of a particular witness is for the trier of fact. It was improper for the Crown to elicit what was, in essence, opinion evidence as to the appellant’s truthfulness from the dispatcher because it is not proper for a witness to give an opinion about the credibility of any other witness: R. v. Marquard, 1993 CanLII 37 (SCC),  4 S.C.R. 223.. R. v. Davis
In R. v. Davis (Ont CA, 2020) the Court of Appeal comments on trial judge reasons on credibility:
 The appellant’s argument invites this court to parse the trial judge’s credibility findings on a word-for-word basis. This is contrary to R. v. Gagnon, 2006 SCC 17,  1 S.C.R. 621, at para. 19, where the court said:. R. v. Slatter
This Court has consistently admonished trial judges to explain their reasons on credibility and reasonable doubt in a way that permits adequate review by an appellate court. Having encouraged these expanded reasons, it would be counterproductive to dissect them minutely in a way that undermines the trial judge’s responsibility for weighing all of the evidence. A trial judge’s language must be reviewed not only with care, but also in context. Most language is amenable to multiple interpretations and characterizations. But appellate review does not call for a word-by-word analysis; rather, it calls for an examination to determine whether the reasons, taken as a whole, reflect reversible error. The task is to assess the overall, common sense meaning, not to parse the individual linguistic components. Although the court went on to say that a court of appeal could not abdicate its responsibility to review the record to see if the findings of fact were reasonably available, it does not follow that the trial judge’s reasons are to be parsed in the manner reflected in the appellant’s submissions. The trial judge listed multiple reasons for not accepting the appellant’s testimony. Read as a whole, those reasons amply support his credibility findings.
In R. v. Slatter (SCC, 2020) the Supreme Court of Canada made useful comments about the testimonial credibility of persons with intellectual or developmental disabilities:
We would simply underline that when assessing the credibility and reliability of testimony given by an individual who has an intellectual or developmental disability, courts should be wary of preferring expert evidence that attributes general characteristics to that individual, rather than focusing on the individual’s veracity and their actual capacities as demonstrated by their ability to perceive, recall and recount the events in issue, in light of the totality of the evidence. Over-reliance on generalities can perpetuate harmful myths and stereotypes about individuals with disabilities, which is inimical to the truth-seeking process, and creates additional barriers for those seeking access to justice. . R v Smith
In R v Smith (Ont CA, 2020) the Court of Appeal considered embellishment of evidence, or the lack thereof, towards credibility:
 The appellant argues that the trial judge erred by finding that McKoy was credible on the basis that she did not embellish her evidence, in violation of this court’s recent decision in R. v. Alisaleh, 2020 ONCA 597. The appellant submits that the trial judge’s finding that she did not embellish her evidence taints his credibility finding and that it cannot stand as a result. We disagree.. Trotter Estate
 It is well established that the fact that a witness does not embellish her testimony does not enhance her credibility. But it is also well established that the mere mention of the absence of embellishment does not undermine a credibility finding that is otherwise properly supported. These points are made clear in Alisaleh, at paras. 16-17, a case in which the Crown conceded that the trial judge had improperly relied on a witness’s lack of embellishment as an “important” factor to “enhance” her credibility, a concession this court accepted.
 This is not a case in which a witness’s lack of embellishment was relied on or used as a “makeweight” to establish or enhance credibility: see R. v. Kiss, 2018 ONCA 184, 145 W.C.B. (2d) 666, at para. 53. It was not inappropriate for the trial judge to note that McKoy had not embellished her evidence in the context of addressing the appellant’s claim that she was merely reiterating press reports or that her testimony was led.
 Nor did the trial judge err by considering McKoy’s testimonial demeanour. His credibility findings were not based on her demeanour; he simply referred to her candour and essential innocence in the manner she gave evidence. He cautioned himself not to rely on demeanour in determining the truth and stated specifically that he was not doing so.
In Trotter Estate (Ont CA, 2014) the Court of Appeal allowed an appeal against the dismissal of a summary judgment motion where the hearing judge dismissed critical evidence without explaining why it's credibility was impugned:
Where important issues turn on credibility, failure to make credibility findings amounts to reversible error: see Sagl v. Cosburn, Griffiths and Brandham Insurance Brokers Ltd., 2009 ONCA 388 (CanLII),  O.J. No. 1879, at paras. 98-100.. R v D.H.
In this criminal case, R v D.H. (Ont CA, 2016), the Court of Appeal discusses the adequacy of a judicial reasons for adverse findings against the credibility of a witness, and their treatment on appeal:
 A trial judge’s assessments of credibility are accorded very considerable deference on appeal, as long as the trial judge has sufficiently explained how significant discrepancies that could undermine credibility and reliability have been resolved. This court recently reiterated that principle in R. v. M. (A.), 2014 ONCA 769 (CanLII), 123 O.R. (3d) 536, at paras. 17-19:
[W]here a case turns largely on determinations of credibility, the sufficiency of reasons must be considered in light of the deference generally afforded to trial judges on credibility findings. It is rare for deficiencies in a trial judge's credibility analysis, as expressed in the reasons for judgment, to warrant appellate intervention: Vuradin, 2013 SCC 38 (CanLII),  2 S.C.R. 639, at para. 11; Dinardo, 2008 SCC 24 (CanLII),  1 S.C.R. 788, at para. 26. While a trial judge is not required to resolve every inconsistency in the evidence, the trial judge “should address and explain how she or he has resolved major inconsistencies in the evidence of material witnesses: G. (M.) (1994), 1994 CanLII 8733 (ON CA), 93 C.C.C. (3d) 347 (Ont. C.A.), at p. 356; R. v. Dinardo, 2008 SCC 24 (CanLII),  1 S.C.R. 788, at para. 31”: M. (A.), at para. 14. As the Supreme Court stated in R. v. Dinardo, at paras. 26-27, the failure to articulate how credibility concerns are resolved, particularly in the face of significant inconsistencies in a complainant’s testimony, may constitute reversible error, as an accused is entitled to know why the trial judge is left with no reasonable doubt.
Nevertheless, the failure of a trial judge to sufficiently articulate how credibility and reliability concerns are resolved may constitute reversible error: Vuradin, at para. 11; Dinardo, at para. 26; R. v. Braich,  1 S.C.R. 903,  S.C.J. No. 29, 2002 SCC 27 (CanLII), at para. 23. After all, an accused is entitled to know why the trial judge had no reasonable doubt about his or her guilt: R. v. Gagnon,  1 S.C.R. 621,  S.C.J. No. 17, 2006 SCC 17 (CanLII), at para. 21.
Similarly, we take it as self-evident that a legal error made in the assessment of credibility may displace the deference usually afforded to a trial judge's credibility assessment and may require appellate intervention.