IS there a meaningful categorical distinction between evidence about the telling of lies, and about the subject matter of those lies, warranting distinct admissibility tests with differing thresholds? This is not just a philosophical conundrum, because of a contentious difference between statutory rules governing the admissibility of sexual behaviour evidence and bad character evidence. The Court of Appeal has wrestled with this question many times, but a black-line distinction has proved elusive, until, perhaps, R. v Hurley [2025] EWCA Crim 642.

The issue emerged soon after Parliament rewrote the admissibility test for evidence of the complainant’s “previous sexual behaviour” or “other sexual experience” (“PSB”) in the Youth Justice and Criminal Evidence Act 1999, s. 41. It is most clearly identified by section 41(4), which deems irrelevant the evidential use of any PSB, the main purpose of which appears to be “impugning the credibility of the complainant as a witness” – notwithstanding that the proposed evidence would otherwise be regarded as relevant, and has successfully navigated one of four gateways (R. v Martin [2004] EWCA Crim 916, [2004] 2 Cr.App.R. 22, at [18]). Parliament failed to address directly the intersection of the new PSB provision with a defence attack on the complainant’s alleged bad character (still a common-law doctrine) including previous false allegations or denials regarding sexual activity.

The judiciary’s immediate concern was that the PSB shield not be circumvented by the same evidence being tendered as bad character. The Court of Appeal ruled that if the defence wished to ask about any alleged previous false complaints as bad character, they also had to seek a ruling that section 41 did not exclude them; to avoid professional misconduct in questioning under the guise of previous false complaints, the defence had to have “a proper evidential basis” for asserting that any statement was (1) made and (2) untrue. If not, the questions would be allocated to section 41 (R. v RT; R. v MH [2001] EWCA Crim 1877, at [41]).

The Criminal Justice Act 2003, s. 100(1) prohibits evidence of “reprehensible behaviour” of a non-defendant witness, with exceptions including where it (1) constitutes important explanatory evidence or (2) has substantial probative value regarding a matter in issue, and is of substantial importance in the context of the whole case. Sexual assault trials are commonly determined by credibility assessments, where the difference between questions going to issue and to credit is reduced to a vanishing point: R. v Funderburk [1990] 1 W.L.R. 587, 597–98 (C.A.). Section 100 makes no reference to section 41, although the difficulty had already been identified in RT.

Hence the conundrum: are alleged lies concerning other sexual activity, including non-consensual, with the defendant or another, about “sexual behaviour”, and caught by section 41, especially the “impugning credibility” prohibition? Or do they only constitute untruthful conduct and therefore channelled through section 100? RT equivocated: “normally” lies were correctly characterised as reprehensible conduct, so evading section 41 (at [33]).

The determinative distinction is often whether there was “a proper evidential basis” or “some material from which it could properly be concluded” that the complaint was false (R. v AM [2009] EWCA Crim 618, at [22]). The court may test the defence’s identification of a permissible purpose for the evidence (e.g. R. v Gabbai [2019] EWCA Crim 2287, [2020] 4 W.L.R. 65, at [59]–[60], [64]). In R. v Wilson [2024] EWCA Crim 1514, at [40], the Court of Appeal risked an element of conflation in importing into section 41(2)(b) (excluding evidence passing through a gateway if unnecessary for a safe conclusion on an issue) the enhanced evidential importance test under section 100(1)(b).

Regardless of the route chosen, the risk remains that admitting the other sexual activity may induce illegitimate lines of reasoning by the jury, which section 41 aspires to intercept (Rook & Ward on Sexual Offences Law and Practice, 7th ed. (2025), at para. 26.217).

In Hurley, on referral from the Criminal Cases Review Commission, Edis L.J. comprehensively reappraised the admissibility of false complaint evidence in sex trials, the first since section 100 was enacted. The Court summarised the current law as follows (at [49]):

The appellant’s submissions reflected biased preconceptions about female complainants. The court’s responses evince an enlightened assessment of the realities often faced by complainants with troubled backgrounds (here, age six when she first became involved with social services):

• • The complainant made three rape allegations against others. She did not report the first to police, as a child of 17, when she understandably had difficulty complaining about her first boyfriend. She abandoned the others, aged 19, as the police blamed her promiscuous lifestyle and alcohol problem (the court deplored this failure to recognise her vulnerability “to sexual predation” (at [60], [85])). She always maintained their truth, so there was no proper evidential basis for falsity.

• • She retracted two rape allegations against her husband as false, in the context of a violent relationship. She was uncertain whether rape was possible within marriage, was emotionally dependent on him in her rape complaint against Hurley (his uncle), and reacted with deteriorating mental health and alcohol self-medication. She told police she lacked the emotional resources for this complaint after Hurley’s prosecution. Contemporary medical records showed she did not want marital sex because of his uncle’s offences. Since consent was disputed, the factual basis for falsity was insufficient, unlike the case against Hurley where he denied any sexual activity (at [79], [86]). Regardless, it “foundered on the rock” of section 41(4), as its sole purpose was to impugn her credibility (at [79]).

• • She did not persist with domestic violence complaints against her husband (engaging only s. 100). The evidence revealed unconvincing retractions and implausible explanations for genuine injuries, with clear indications of repeated abuse, and credible explanations for dropping the allegations.

The Court castigated “as a dangerous line of logic” the defence position that where a person made multiple rape allegations, that in itself could generate an inference of falsity (at [90]): every case required contextualised consideration of the facts. The CCRC had concluded that the complainant was deploying a “self-serving” strategy in maintaining her allegations against Hurley, to avoid repercussions including “criminal liability” for making other false complaints. This was emphatically rejected by the Court of Appeal as “unconvincing” and “fanciful” (at [109]).

Welcome as these judicial signals are, point (iii) of the court’s legal summary shows that the tension between previous sexual behaviour and credibility in section 41 and section 100 remains unresolved. Canadian courts take a more direct route to a satisfactory resolution. Their position is that the “other sexual activity” evidence shield in the Criminal Code of Canada, s. 276 does not apply to any prior complaint of non-consensual sexual activity, because there is no risk of invoking the twin myths proscribed in section 276(1) (that someone sexually experienced would be more likely to consent, and in any case is less worthy of belief): R. v Seaboyer [1991] 2 S.C.R. 577, at [59]. Instead, the common-law admissibility rules apply, focusing on relevance to a specific fact in issue, and whether its prejudicial effect substantially outweighs its probative value (R. v B(O) (1995) 45 C.R. (4th) 68 (NSCA), at [61]). This approach sharply focuses judicial attention on the real hazards of such evidence for the decision maker’s deliberations, which can be overlooked in England’s current category approach. An additional Canadian safeguard is that much higher proof of falsity is required for admission of similar allegations of sexual assault against third parties under the common-law doctrine. The defence must establish either that the complainant had recanted the earlier accusation, or that it was “demonstrably false”: R. v Riley (1992) O.R.(3d) 151 (Ont CA), 154; R. v B(AR) (1998) 41 O.R.(3rd) 361 (Ont CA), at [15], affd. [2000] 1 S.C.R. 78.

The Law Commission in Evidence in Sexual Offences Prosecutions: a Final Report (Law Com No 420) (HC 1156, July 2025) has not embraced this solution, instead preferring to preserve the present unworkable distinctions. It would add “myths and misconceptions” as a statutory consideration to section 100 (apparently even where sex is not involved) (paras. 4.37–4.38, Recommendation 21). Both provisions would be augmented by a “substantial probative value” threshold. The Law Commission abandoned, without explanation, its provisional proposal that previous false allegations be parked under section 41 with its more stringent tests. This final recommendation would only perpetuate artificial distinctions between the various forms and subject matter which lies about sex may take, instead of weighing the proof of alleged falsity, and its probative value and prejudicial impact for the decision maker’s reasoning (the Commission’s formulation of “myths and misconceptions” being only one, as Hurley demonstrates). The Commission’s recommendation, if implemented, would only exacerbate the issues lingering after Hurley, embedded in flawed legislation focused on delineating categories, without principled reasons for different analytical frameworks, rather than on risks to the verdict’s integrity from distorted reasoning, arising beyond the ambit of section 41. In the meantime, Hurley has banished many biased considerations from this tricky analysis of two statutes irreconcilable in principle and application.