Threshold for Leaving Self-Defence to the Jury: No Direction Where the Evidential Basis Would Invite Speculation
Case: R v Hussain [2026] EWCA Crim 699
Court: England and Wales Court of Appeal (Criminal Division) (Lady Justice May DBE, Mr Justice Linden, Ms Justice Obi DBE)
Date: 29 April 2026
Appeal from: Crown Court at Luton (His Honour Judge Kay KC)
1) Introduction
The appellant was convicted of murder following a jury trial and sentenced to life imprisonment with a 21-year minimum term. He was acquitted of possessing a bladed article contrary to section 139 of the Criminal Justice Act 1988. The central issue on appeal was narrow but important: whether the trial judge was required to direct the jury on self-defence (and therefore leave self-defence as a route to acquittal), notwithstanding that the appellant’s case was that he did not stab the deceased and that an unknown third party may have done so.
The prosecution case relied heavily on CCTV evidence showing the appellant’s movements before and after the stabbing, his interaction with his waist bag/waistband, and his rapid departure, together with subsequent steps to evade arrest. The appellant’s evidence was that the deceased struck him with a walking stick inside the block; he pushed the deceased and ran, noticing a “mystery man” at the end of the hallway. He denied stabbing the deceased and denied bringing a weapon.
2) Summary of the Judgment
The Court of Appeal dismissed the appeal and held that the conviction was safe. While reaffirming that self-defence may need to be left even if not formally relied upon (including where the defence is alibi), the Court held that on the facts here there was no evidential basis capable of amounting to a prima facie case of self-defence. Leaving self-defence would have invited speculation—particularly as the appellant’s own account did not support that stabbing with a bladed article was a necessary or proportionate response to any perceived threat.
3) Analysis
3.1 Precedents Cited
(a) R v Bonnick (1978) 66 Cr App R 266
The Court relied on Bonnick for the governing threshold: self-defence should be left “when there is evidence sufficiently strong to raise a prima facie case of self-defence if it is accepted”; otherwise, a direction would “invite speculation”. The judgment emphasises that this is a common-sense, evidence-specific evaluation for the trial judge.
“Self-defence should be left to the jury when there is evidence sufficiently strong to raise a prima facie case of self-defence if it is accepted. To invite the jury to consider self-defence upon evidence which does not reach this standard would be to invite speculation.”
(Bonnick, quoted at [27])
The Court treated Bailey as confirming two propositions:
- hopeless defences with “no factual basis of support” need not be left; but
- if the accused’s account includes matters that, if accepted, could raise a prima facie case of self-defence, it should be left even if not “formally relied upon”.
Importantly, Bailey illustrates that a self-defence direction may be required even where the defendant advances another explanation (there, accident), because rejection of that explanation could logically open a route to self-defence on the evidence. The Court in Hussain distinguished that logic on the facts: even accepting the appellant’s evidence here, the necessary and proportionate use of a bladed weapon did not arise as a realistic inference.
Mula demonstrates that self-defence can be raised by evidence from a complainant even where the defendant relies on alibi. The Court noted that in Mula there was evidence consistent with the defendant believing he was about to be struck, making self-defence a live issue. In Hussain, the Court accepted that the absence of an explicit assertion of fear is not determinative, but held that the remaining evidence still did not cross the prima facie threshold.
(d) R v Harewood and Rehman [2021] EWCA Crim 1936
The Court invoked Harewood and Rehman to underline an appellate restraint point: trial judges often have a better “feel” for whether a defence is truly in issue, especially where the appeal court lacks a transcript of all evidence. This did not replace the legal test, but supported deference to the trial judge’s evaluative assessment at the margin.
3.2 Legal Reasoning
(a) The controlling test: prima facie self-defence vs speculation
The Court reaffirmed the orthodox approach: a judge must leave self-defence where there is evidence which, if accepted, could amount to self-defence; but need not leave it where the evidential basis is too thin and would invite jury speculation.
(b) “No express fear” is not a legal bar—but may be evidentially significant
The Court accepted that the trial judge was wrong if he had treated the absence of an express claim of fear/necessity as automatically fatal. But it concluded that the trial judge did not apply such a rigid rule. Rather, the judge was evaluating the evidence in context and concluding that there was no proper evidential platform for the two key components of self-defence.
(c) Why the evidence did not raise self-defence on these facts
The Court’s reasoning turned on the mismatch between (i) the asserted threat and (ii) the hypothesised response (a stabbing to the neck with at least moderate force):
- Equivocal pre-entry aggression: Whatever earlier tension existed outside appeared to have dissipated before both men entered the building. The CCTV suggested a calm transition into the communal area.
- Post-stabbing “chase” was ambiguous: CCTV showed the deceased following the appellant out, but by then the deceased had already been stabbed; this was not reliable evidence of a pre-stabbing threat justifying force.
- The appellant’s own account did not support necessary/proportionate stabbing: Even accepting the appellant’s claim that he was struck with a stick, the Court held that, objectively, the reasonable response on that account was to back off or leave—something the appellant in fact did.
- Relative capacities: The appellant was younger and faster; the deceased was older and used a stick. Those factors mattered to the plausibility of stabbing as a necessary and proportionate defensive act on the evidence presented.
The Court therefore held that any self-defence narrative would require the jury to speculate about additional facts not in evidence—precisely what Bonnick warns against.
(d) Manslaughter left, self-defence not left: no inconsistency
The Court rejected the argument that leaving manslaughter (as “borderline”) required leaving self-defence. The judge left manslaughter on the distinct question of intent (whether there was a viable basis to doubt intent to kill or cause really serious harm). That did not make the use of a bladed weapon in the neck potentially reasonable or proportionate as a defensive response.
The Court added a practical safety point: the jury’s murder verdict meant they were sure at least of intent to cause really serious harm. Given that finding and the evidential picture, it was “vanishingly unlikely” that self-defence would have succeeded even if left—reinforcing the conclusion that the conviction was safe.
(e) Prosecution burden to disprove self-defence arises only if self-defence is in issue
The appellant argued that unlawfulness was not conceded and so the prosecution had to disprove self-defence. The Court clarified that the prosecution’s obligation to negative self-defence arises only once a prima facie case is raised on the evidence. As none was raised here, the judge was not required to direct on self-defence.
3.3 Impact
- Reinforcement of the “anti-speculation” gatekeeping function: The decision strengthens the message that a self-defence direction is not automatic whenever there is some evidence of hostility or violence; there must be an evidential foundation capable of supporting both necessity (as perceived) and proportionate force.
- Clarity in “I didn’t do it” cases: The judgment illustrates that even where the defendant accepts presence and alleges a third-party assailant, self-defence will only be left if the evidence (from any source) can realistically support it without conjecture.
- Appellate caution and trial-judge vantage point: By relying on R v Harewood and Rehman [2021] EWCA Crim 1936, the Court signals that appellate intervention is less likely where the decision is a fact-sensitive evaluative judgment and the full evidential texture is better appreciated at trial.
4) Complex Concepts Simplified
(a) What does “leave self-defence to the jury” mean?
It means the judge must give the jury legal directions on self-defence and allow them to acquit if they accept it (or if it raises a reasonable doubt), instead of treating guilt as turning only on whether the defendant committed the act and had the required intent.
(b) What is a “prima facie case” of self-defence?
It is a minimum evidential threshold: if the jury accepted the supporting evidence at its highest, self-defence could succeed in law. If the evidence cannot support self-defence without filling gaps by guesswork, it is not prima facie.
(c) Why is “speculation” a problem?
Criminal verdicts must be based on evidence. A judge should not invite jurors to invent missing facts (for example, about what threat existed or what the defendant believed) to make a defence fit.
(d) Relationship between intent (murder vs manslaughter) and self-defence
Intent addresses what the defendant meant to do. Self-defence addresses whether the force used was lawful because it was necessary (as perceived) and reasonable/proportionate. A case may raise doubt about intent (supporting manslaughter) without raising any realistic basis for lawful force (self-defence).
5) Conclusion
R v Hussain [2026] EWCA Crim 699 confirms that self-defence must be left to a jury only where the evidence, taken at its highest, raises a genuine prima facie basis for it. The Court emphasised that, even accepting the appellant’s account of being struck, nothing in the evidence could realistically justify a neck stabbing with a bladed implement as a necessary and proportionate response; leaving self-defence would therefore have invited impermissible speculation. The decision is a reminder that the duty to direct on self-defence is principled but not automatic, and that the trial judge’s evaluative assessment of the evidential foundation will often be upheld on appeal.