3.1 Precedents Cited

R v Stott [2018] EWCA Crim 538

The court treated Stott as a close comparator: where material is properly admissible as part of the narrative/history relevant to issues such as consent, and the summing-up keeps the jury focused on that issue, appellate intervention is unlikely. The court noted the trial judge’s approach here was “very similar” to that in Stott.

R v Sule [2013] Cr App R 3

Cited alongside Stott for the proposition that “background” relationship material may fall within s.98 (i.e., not “bad character” requiring a s.101 gateway), because it is part of the relevant narrative rather than propensity evidence. This underpinned the court’s conclusion that the challenged material was properly admissible under s.98.

R v Hunter [2015] EWCA Crim 631; [2015] 1 WLR 5367

Hunter provided the controlling appellate methodology. The court emphasised:

• appeals on directions (including “character” directions) are highly context-sensitive;
• trial judges have a “feel for the case”;
• where counsel did not take the point at trial (and here counsel effectively endorsed the draft directions), that is a strong indicator no material unfairness occurred.

This was critical to rejecting the submission that the absence of additional limiting directions rendered convictions unsafe.

R v MA [2019] EWCA Crim 178

The appellant relied on MA to argue that the jury should have been directed on how to use controlling/abusive relationship evidence admitted under s.98. The court distinguished MA on its facts: it involved prominent evidence of repeated physical non-sexual violence not charged on the indictment, which assumed major prominence at trial, creating a risk of impermissible reasoning unless carefully directed. By contrast, here the s.98 material was (i) largely sexual relationship history and control relevant to consent and (ii) did not introduce an additional, prominent “violence narrative” beyond the indicted sexual offences.

R v Hamadi [2007] EWCA Crim 3048

Hamadi was applied in the s.41 context, particularly on the requirement that proposed sexual-behaviour evidence must be truly probative of a live issue and not a disguised attempt to attack credibility. The trial judge’s categorical assessment of the “B” messages was aligned with Hamadi, and the Court of Appeal treated those reasons as “unimpeachable”.

R v Wilson [2024] EWCA Crim 1514; [2025] 1 WLR 2865

Wilson reinforced that s.41 rulings—when the judge correctly self-directs on the law—are quintessentially evaluative and will rarely be overturned. The court used Wilson to frame ground 5 as an attempt to re-argue an exercise of judgment rather than identify legal error.

3.2 Legal Reasoning

(A) s.98 Criminal Justice Act 2003: “Bad character” versus “relevant background”

The court proceeded on a clear organising distinction:

• s.98 material: evidence of misconduct that is not “bad character evidence” for the purposes of the Act—typically because it is part of the relevant narrative/context (here, the relationship history and control relevant to consent).
• s.101 material: evidence that is bad character and requires a statutory gateway (e.g., important explanatory evidence under s.101(1)(c)).

Critically, by the time of the appeal the appellant accepted admissibility under s.98 and did not pursue challenges to admission (grounds 1 and 3 fell away). The appeal therefore narrowed to whether the directions were defective.

(B) Whether additional “limiting use” directions were required

The court assessed the summing-up as a whole. It held the judge’s directions:

• correctly focused the jury on the live consent question, including the key distinction: “submission is not consent”;
• squarely located the relationship-history evidence (including the alleged coercive dynamic) as relevant to whether C was worn down into submission or instead gave reluctant but real consent;
• addressed the defence “safe word / consensual non-consensual” narrative, directing the jury how prior agreement (if found) could affect both actual consent and reasonable belief in consent.

The court rejected the argument that a separate, formulaic direction was required simply because the evidence involved misconduct (stalking) or uncharged behaviour in the ABE material. The crucial question was functional: did the directions, in context, prevent impermissible propensity reasoning and keep the jury anchored to the permitted relevance? On these facts, they did.

(C) The stalking conviction

The stalking (via a monitoring app) was treated as playing only a small role at trial, and it was presented as evidence of the nature of the relationship—control/possessiveness—rather than as a basis to infer sexual offending propensity. The court noted:

• no suggestion the prosecution invited improper use of the stalking evidence;
• the judge’s summing-up confined it to relationship context;
• the appellant also benefited from an agreed “qualified” good character direction (no prior sexual convictions), further reducing the risk of unfair prejudice.

(D) Counsel conduct and draft directions: the Hunter factor

The court attached weight (per R v Hunter [2015] EWCA Crim 631; [2015] 1 WLR 5367) to the contemporaneous reality: draft directions were circulated; defence counsel replied that they were “fine” (save typographical points); no recorded request was made for the additional directions later advanced on appeal. This strongly supported the conclusion that the trial was experienced as fair and the directions were adequate.

(E) Safety proviso in substance

The court added that even if additional direction had been desirable, the convictions would remain safe given the strength of other evidence—especially the videos made by the appellant and explicit text messages—supporting the prosecution case on consent and knowledge.

(F) s.41 YJCEA 1999: sexual behaviour evidence and fair trial balance

The judge’s staged approach to “B” messages was upheld:

• careful categorisation of the proposed material;
• exclusion where insufficient similarity/probative value existed or where the true purpose was credibility attack;
• on the renewed application (after C denied fantasies/role play), a tailored remedy under the fair-trial lens: permit a limited question to prevent the jury being misled, escalating to adducing the messages only if C denied sending them.

The Court of Appeal treated this as a textbook exercise of discretion consistent with R v Hamadi [2007] EWCA Crim 3048 and R v Wilson [2024] EWCA Crim 1514; [2025] 1 WLR 2865.

(G) Sentence: severe psychological harm without expert evidence

The court reaffirmed that severe psychological harm can be found without expert evidence, including by reference to judicial observation of the complainant giving evidence and to a factually grounded victim personal statement. The categorisation of the lead rape (count 3) as Category A2 (starting point 10 years) was upheld, and the overall 12-year totality outcome was “severe” but not “manifestly excessive”.

3.3 Impact

(A) Directions on s.98 material: no automatic “bad character-style” warning

The practical significance of R v Sands is its confirmation that where:

• relationship-history/control evidence is admitted under s.98,
• its relevance is directly tied to the live issue (here, consent/submission and reasonable belief), and
• the summing-up as a whole keeps the jury confined to that relevance,

the trial judge will not necessarily be required to provide additional bespoke directions of the type often given with s.101 “bad character” propensity evidence. The decision also underlines the appellate reluctance (per Hunter) to micromanage summings-up, particularly where trial counsel endorsed draft directions.

(B) s.41: proportionate, issue-specific admission remains the model

The case supports a “minimal necessary” approach to s.41: permit enough to avoid misleading the jury and to ensure fairness, but not so much as to reintroduce the very stereotype-driven reasoning s.41 is designed to prevent. In practice, “gateway” arguments based on similarity must show true probative force on a live issue, not merely contextual salience.

(C) Sentencing: evidencing psychological harm

Sands adds to the steady line of authority that sentencing judges may find severe psychological harm by judicial observation and credible lay material, without requiring expert reports in every case—particularly where the consequences are longstanding and plainly articulated.