THE DUTY DOES NOT FLICKER ON AND OFF – ZZZ v Ministry of Defence [2026] EWCA Civ 875: occupiers’ liability, low balustrades, horseplay and the discipline of factual analysis

July 13, 2026

By Pankaj Madan

The Court of Appeal has upheld a finding of full liability against the Ministry of Defence after an 18-year-old soldier sustained a significant brain injury when he fell over a low landing balustrade in barracks. In my view, the judgment is important well beyond its military setting. It addresses the proper limits of appellate intervention, the status of a visitor under the Occupiers’ Liability Act 1957, the relevance of current standards to older premises, causation in falls from height, and the evidential burden on a defendant alleging contributory negligence.

Introduction

Falls on or from stairs are often treated as intensely fact-sensitive cases with limited wider significance. That is usually a mistake. Such cases frequently bring into sharp focus the central questions in occupiers’ liability: what was the danger; who ought reasonably to have appreciated it; what precautions were proportionate; and did the condition of the premises materially contribute to the injury?

Where the injury is a brain injury, the evidential difficulties become more acute. Memory may be fragmented or unreliable. Early histories may differ. Accounts may evolve as a claimant attempts to reconstruct events which he or she cannot properly remember. The court must be careful not to equate inconsistency caused by injury with dishonesty.

ZZZ v Ministry of Defence [2026] EWCA Civ 875 is a striking illustration. The claimant, an 18-year-old rifleman required to live in Army accommodation, fell from a second-floor landing to the ground floor following an alcohol-exacerbated episode of horseplay. The balustrade was approximately 900 mm high. The claimant was 6 ft 4 in tall. He sustained serious injuries, including a significant brain injury. At first instance, HHJ Simon, sitting as a Judge of the High Court, in July 2025, found the Ministry of Defence liable in full. On 8^th July 2026, The Court of Appeal, in a judgment given by Stuart-Smith LJ, with which King and Falk LJJ agreed, dismissed all six grounds of appeal.

The judgment applies established principles with clarity. Its central message is that an occupier cannot avoid responsibility for a dangerous feature of premises by attempting to convert a resident into a momentary trespasser whenever the resident behaves imperfectly. Nor can an appeal against evaluative findings be turned into a second trial by selecting isolated inconsistencies from a much wider evidential picture.

The accident and the decision at first instance

The claimant was based at Bulford Army Base and was required to live in the barracks. In the early hours of 24 March 2019, after drinking with other young soldiers in the barracks and in Salisbury, he and Rifleman Graham became involved in a struggle on the landing.

The precise mechanism of the fall was disputed. The claimant’s accounts varied, unsurprisingly given the combination of a serious brain injury and alcohol. Statements taken from fellow soldiers described playfighting or a “friendly drunk fight”, involving grappling, a headlock and an attempt by Graham to break free. The claimant came into contact with the balustrade, pivoted over it and fell to the ground floor.

The building dated from about 1939 and had originally been constructed as a place to treat the sick rather than for use as soldiers’ accommodation. The balustrade was approximately 900 mm high. Modern guidance for guarding to common landings indicated a height of 1100 mm. It was common ground between the experts that the Building Regulations were not retrospective and had not been triggered by the refurbishment works. That did not, however, answer the common law or statutory question whether the premises were reasonably safe for their actual use.

HHJ Simon found that the incident occurred during a “frolic”, by which he meant something akin to playfighting exacerbated by alcohol, rather than a malicious assault. He regarded the claimant as credible, in the sense that he was doing his best to tell the truth, but unreliable in his recollection of significant events. He preferred the broad account emerging from the contemporary witness statements.

The judge found that the claimant had not ceased to be a lawful visitor; that the pre-accident risk assessment was inadequate; that the low balustrade was an inherently unsafe feature; and that a balustrade of around 1100 mm would probably have prevented the fall. He rejected the defences of volenti, illegality and contributory negligence.

The appeal

The Ministry of Defence advanced six grounds of appeal. Taken together, they amounted to a sustained attack upon both the factual findings and the resulting conclusions of law.

The Ministry contended that:

The judge’s findings that the event was a frolic and involved no great force were unsupported by the evidence.
The claimant had breached Army rules or committed criminal offences and had thereby ceased to be a lawful visitor.
The claim was barred by illegality.
The judge had wrongly treated the balustrade as a danger, applying modern standards retrospectively and reasoning with hindsight.
Causation had not been established because the claimant’s expert could not exclude the possibility that the claimant would have fallen over a 1100 mm balustrade if subjected to a forceful push.
The judge ought to have made a substantial deduction for contributory negligence, with the Ministry contending for a deduction of 75 per cent.

Every ground failed.

An appeal is not a re-run of the trial

The first ground challenged the trial judge’s findings about what occurred on the landing. The Court of Appeal began with the restraint required of an appellate court when reviewing findings of fact and evaluative conclusions.

Drawing upon Fage UK Ltd v Chobani UK Ltd [2014] EWCA Civ 5 and Prescott v Potamianos [2019] EWCA Civ 932, Stuart-Smith LJ emphasised that the appellate court does not undertake the factual balancing exercise afresh. The issue is not whether the appellate judges might themselves have reached a different view. The question is whether the trial judge’s decision was undermined by an identifiable flaw, such as a gap in logic, an inconsistency or a failure to take account of a material factor.

That discipline was decisive. There were features of the evidence capable of supporting different findings, but that was not enough. There was evidence from Riflemen Green and Peden supporting the judge’s conclusion that the incident remained playfighting and did not involve a concerted attempt to throw the claimant over the balustrade.

Green described the two men as laughing and giggling. He described Graham breaking out of a headlock, apparently striking the claimant’s arm during the movement, whereupon the claimant fell backwards against the balustrade and flipped over it. Peden described a friendly drunken fight involving grappling and shoving. He believed that Graham, whose face was pressed against the claimant’s chest, would not have known that they were moving towards the balustrade.

The Court of Appeal described the case as a classic example of an incident about which the evidence was inconsistent and in respect of which it was the primary function of the trial judge to make the necessary assessments. The judge had paid close attention to the evidence and recognised the inconsistencies. Although different findings might have been available, the Court was not compelled to overturn those which he made, “not by a long chalk”.

That is a useful reminder of the distinction between showing that another factual conclusion was possible and demonstrating that the conclusion actually reached was wrong. The former is not sufficient on appeal.

The use of hearsay evidence

The judgment contains an important procedural lesson about hearsay evidence.

The Ministry of Defence served statements made by the soldiers to the service police but did not call the relevant witnesses. The claimant obtained an order permitting him to call three of the soldiers for cross-examination and asked the Ministry for their contact details. The response was that two had left the Army and the location of the third was uncertain.

Evidence at trial indicated that serving soldiers could be traced where their regiment and rank were known, and that the Army retained discharge addresses for those who had left. The Court of Appeal considered that there was no apparent reason why the requested information could not have been provided. It described the Ministry’s response as a “brush off” which had two malign effects: the claimant could not cross-examine the witnesses, and the judge was left to assess the untested statements which the defendant itself had introduced.

That background considerably weakened the Ministry’s attempt on appeal to subject those statements to a minute forensic examination in search of inconsistencies. A party which chooses to place hearsay evidence before the court, but does not call the witnesses or facilitate their attendance, may face real difficulty in subsequently inviting the court to reject the same evidence as unreliable.

There was a further point of procedural importance. The trial judge’s findings did not precisely reflect either party’s pleaded case. The Court of Appeal considered that best practice would have been for the judge to invite further submissions once he had formulated the findings he was minded to make. The failure to do so did not justify setting the judgment aside, because the findings were supported by the evidence and the exhaustive submissions on appeal came nowhere near demonstrating that they should not have been made. Nevertheless, where a judge is considering a factual route which differs materially from the competing pleaded cases, procedural fairness may require the parties to be heard.

Credibility, reliability and brain injury

The treatment of the claimant’s evidence is particularly important in catastrophic injury litigation.

The claimant had given a number of differing accounts. At various stages, he described missing a step, falling backwards, being lifted, having his legs pulled from beneath him, being picked up and thrown, being pushed from behind and being pushed in the chest. His orientation to the balustrade also varied between accounts.

These inconsistencies entitled the trial judge to conclude that the claimant was not a reliable witness as to the mechanism of his fall. They did not require a finding that he was dishonest. The judge found that the claimant was doing his best to give truthful evidence but that his recollection had been affected by his injuries and alcohol.

That distinction between credibility and reliability was both humane and legally sound. A witness may be entirely honest yet unable to provide a dependable chronology because of post-traumatic amnesia, cognitive impairment, intoxication, medication or the ordinary process of memory reconstruction.

The Court of Appeal also noted that it had been submitted that gaps and inconsistencies in the claimant’s evidence were “convenient”. That allegation had not been put to the claimant in cross-examination. If a party intends to suggest that a claimant is deliberately exploiting gaps in memory or has fabricated an account, fairness requires the allegation to be put squarely.

Early medical histories should be preserved and analysed carefully, but they should not be treated as though each were a formal witness statement. Medical notes may contain abbreviated histories taken in difficult circumstances from a severely injured, medicated, confused or cognitively impaired patient. Their value depends upon the circumstances in which they were created.

The proper task is to construct the mechanism from the whole evidential mosaic: eyewitness accounts, photographs, scene measurements, contemporary accident documents, engineering evidence and the medical explanation for impaired memory. The claimant’s recollection is important, but it need not carry the whole case.

A lawful visitor does not become a transient trespasser

The most significant part of the judgment concerns the claimant’s status under the Occupiers’ Liability Act 1957.

The Ministry argued that the claimant’s conduct contravened Army rules or amounted to criminal conduct, taking him outside his express or implied licence to use the premises. It alleged potential offences under the Armed Forces Act 2006 and affray under the Public Order Act 1986.

The Court rejected both the factual premise and the proposed legal principle.

On the findings made at first instance, the claimant was responding to horseplay initiated by Graham. The headlock was consistent with the claimant’s military training in de-escalating potentially aggressive behaviour. There was no proper basis for finding that he had committed a disciplinary or criminal offence. There had been an investigation, but no prosecution or disciplinary proceedings followed.

More fundamentally, the claimant was on the second floor because his employer required him to live there. The landing formed part of the common areas of his accommodation.

The principle advanced by the Ministry was far too broad. It appeared to suggest that a lawful visitor who engaged in conduct which the occupier would require him to stop thereby became a trespasser. Stuart-Smith LJ rejected the suggestion that James v White Lion Hotel [2021] EWCA Civ 31 supported such a general rule.

The relevant question under the 1957 Act is whether the person is a visitor, meaning someone who enters or uses the premises with the occupier’s express or implied permission. Conduct of which the occupier might disapprove does not, without more, transform a lawful visitor into a trespasser.

The Ministry’s case also created an insuperable practical difficulty: in relation to precisely what area had the claimant become a trespasser?

He could not sensibly have become a trespasser in the barracks as a whole because he was required to live there. He could not have been a trespasser in his own room. The argument therefore reduced itself to the proposition that he became a trespasser on the landing for the brief period during which he responded to Graham’s renewed horseplay, only to become a lawful visitor again when the interaction ended.

The Court described that as a startling and unworkable framework. It would mean that the Ministry owed no duty concerning the balustrade during the brief struggle, but would have owed a duty if the claimant had returned the following morning and fallen after stumbling. Stuart-Smith LJ did not accept that the common duty under the 1957 Act was so “ephemeral”.

This distinction is of considerable practical importance. Courts and practitioners should keep separate the following questions:

Was the claimant a visitor?
Did the occupier take such care as was reasonable to see that the claimant was reasonably safe in using the premises for the permitted purpose?
Did the claimant fail to take reasonable care for his own safety?

Misconduct may be relevant to breach, causation or contributory negligence. It does not necessarily extinguish the underlying duty by retrospectively converting the claimant into a trespasser.

The relevance of James v White Lion Hotel

The Court’s reliance upon James v White Lion Hotel is particularly apposite.

In that case, the Court of Appeal recognised that the common duty of care is owed by a particular occupier to a particular visitor in a particular factual setting. A hotel room is a “home from home”. A hotel guest may be tired, off guard, relaxed or affected by alcohol. Those are facts of life for a hotelier and form part of the context in which the safety of the premises must be assessed.

There was a clear parallel with compulsory military accommodation. The claimant was not engaged in an inherently dangerous recreational activity of the kind considered in Tomlinson v Congleton Borough Council. He was in the common area of the place in which he was required to live.

Army witnesses accepted that young soldiers drinking in barracks, going out drinking and engaging in frolic or playfighting were familiar occurrences. One witness described such behaviour as having existed for as long as the Army itself. The King’s Regulations also recognised service accommodation as personnel’s own space away from working routines, while requiring the Army to continue fulfilling its obligations as a responsible employer.

This did not mean that the Ministry was required to prevent every act of horseplay. It meant that the foreseeable behaviour of those living in the building formed part of the factual context in which the adequacy of the guarding had to be assessed.

Old construction is not a complete answer to present danger

The Court of Appeal accepted that the Building Regulations were not retrospective. It did not treat modern guidance as imposing a direct statutory obligation to reconstruct every balustrade in an older building.

Instead, it treated modern regulations and standards as evidence of accumulated professional knowledge and practical experience about sensible safety precautions. The Building Regulations represented the consensus of current professional opinion about appropriate design and construction. Conversely, compliance with standards applicable at the date of construction did not mean that a building could never subsequently be assessed as presenting an unacceptable risk.

That is the correct analytical route. Historic compliance is relevant, but it is not conclusive forever. An employer or occupier must keep safety under review by reference to:

the present use of the building;
the characteristics of its users;
changes in professional knowledge;
the likelihood and gravity of injury; and
the availability and cost of precautions.

A feature which was lawful when constructed may nevertheless present an unacceptable risk many years later, particularly where the use of the premises has changed, the exposed population is different, or a straightforward and inexpensive improvement is available.

The evidence supporting the finding that the balustrade was dangerous was formidable.

The balustrade was approximately 900 mm high, around 200 mm below modern guidance for common landings. Photographs ( see above taken from the case report) demonstrated its height relative to the claimant, who was tall but, as the Court noted, not “freakishly tall”. The Army’s own witness accepted that the risk shown by the photograph was obvious. Alcohol and horseplay among young soldiers were known features of life in the accommodation.

The post-accident risk assessment categorised the likelihood of a fall as “common, regular or frequent occurrence” and the potential severity as including fatalities or major injury. The Ministry’s learning account concluded that the balustrade had not been high enough to prevent the fall. A subsequent Authority Notice of Change recorded that low balustrades across the retained estate posed a severe risk of occupants falling over them.

The relevant remedial works were modest. Mr Watts described raising the balustrade as achievable at minimal cost. The Court of Appeal endorsed that description when considered against the catastrophic potential consequences of a fall.

The balancing exercise was therefore stark. The foreseeable consequences were death or catastrophic injury. The cost and difficulty of the precaution were modest. The exposed population included young soldiers whose drinking and horseplay were foreseeable. The existing balustrade was materially lower than the height accepted in modern professional guidance.

Hindsight and inadequate risk assessment

The Court’s treatment of hindsight is particularly valuable.

It was true that nobody had identified the balustrade as dangerous before the accident. That was not determinative. The correct question was not whether the Ministry did recognise the danger, but whether it should have recognised it.

Previous risk assessments had failed to identify the danger. That did not establish that the balustrade was safe. It supported the trial judge’s conclusion that those assessments had been inadequate. Repeated inadequate inspections do not convert an unsafe feature into a safe one.

The duty was anticipatory and continuing. An occupier or employer cannot necessarily wait for a near miss or serious accident before undertaking a sufficiently robust assessment. The purpose of risk assessment is to identify foreseeable risks before injury occurs.

This is especially important in stair and balustrade claims. A low guarding feature may remain in place for many years without an accident. That history may be relevant, but it is not conclusive. The correct inquiry remains whether the risk was reasonably foreseeable, whether the possible consequences were serious, and whether proportionate precautions were available.

Causation must be tied to the facts actually found

The causation appeal illustrates a recurring error in the treatment of expert evidence.

The Ministry relied upon a concession made by the claimant’s engineer, Mr Watts. He accepted in cross-examination that he could not say whether an additional 200 mm of balustrade height would have prevented a fall if one person pushed another with all the force he could muster in an effort to knock him from his feet.

The difficulty was that the factual premise of the question had not been accepted by the judge. The judge found no great force, no lifting of the claimant and no concerted attempt by Graham to propel him over the balustrade.

On the facts actually found, Mr Watts’ evidence remained that a man of the claimant’s height falling against a 900 mm balustrade faced a relatively high possibility of toppling over it, whereas he was unlikely to topple over a 1100 mm balustrade. The judge was entitled to accept that evidence and to conclude that the fall would probably have been avoided had the balustrade been raised.

Expert concessions are only as useful as their underlying factual assumptions. A concession based upon an extreme or rejected version of events does not displace the expert’s opinion on the factual mechanism which the court accepts.

In stair and balustrade cases, the expert evidence should therefore be mapped against each realistic factual scenario. The expert should be asked:

What is the probable mechanism of the fall on each available account?
What part did the height, geometry and position of the guarding play?
What would probably have occurred had the guarding been at a reasonably safe height?
Does the answer change according to whether the claimant stumbled, was lightly displaced, was pushed with moderate force, or was subjected to a deliberate and forceful propulsion?
Which conclusions can be expressed on the balance of probabilities, and which amount only to a reduction in risk?

The question is not merely whether safer guarding would have reduced the risk in some general sense. The claimant must prove that, on the facts found, the precaution would probably have prevented the accident or injury.

Contributory negligence: difficult, but not proved

The sixth ground was the closest. Stuart-Smith LJ expressly acknowledged that he had not found it easy to determine.

The Ministry argued that fighting or playfighting near an open stairwell was objectively reckless. It submitted that the claimant had voluntarily returned to the landing, could have walked away and had failed to take reasonable care for his own safety. It contended for a reduction of 75 per cent.

There was an immediate attraction in the argument at a general level. Horseplay close to an open stairwell may plainly create a risk of serious injury. However, contributory negligence is not determined at a level of generality. It depends upon the facts established in relation to the particular claimant.

Several matters were important.

First, the claimant’s evidence could not safely be given substantial weight where the judge had found him unreliable as to significant events.

Secondly, the Ministry’s characterisation of the conduct did not sit comfortably with the judge’s findings. The interaction had been re-instigated by Graham. There was no serious intent to fight. The fall was accidental, not malicious.

Thirdly, the claimant’s use of the headlock was consistent with his training in de-escalation.

Fourthly, there was no precise finding about the duration of the interaction. The available evidence suggested that the relevant sequence may have been swift and short.

Fifthly, the Court did not consider that the claimant should be criticised for failing to appreciate the unusual danger created by the low balustrade when the Ministry’s own inadequate risk assessments had failed to identify it.

Sixthly, the landing was relatively extensive and there were no sufficiently precise findings about where the interaction began or the parties’ exact position before the claimant went over the balustrade.

The burden of proving contributory negligence rested upon the Ministry. The Court did not have sufficient confidence in the fine detail of the incident to reverse the trial judge and make a deduction.

The result should not be read as establishing that horseplay near stairs can never amount to contributory negligence. It plainly can. The decision is instead a reminder that apportionment must be proved rather than assumed.

General disapproval of the claimant’s conduct is not a substitute for evidence and findings addressing:

what the claimant actually did;
what danger he knew or ought reasonably to have appreciated;
whether he had a meaningful opportunity to avoid or disengage from the risk;
how long the conduct continued;
whether his actions made a causative contribution to the fall; and
the relative blameworthiness and causative potency of the parties’ conduct.

Illegality and volenti

The illegality ground failed once the factual allegations of criminal conduct failed.

The Court of Appeal added that it would have dismissed this ground even if criminal conduct by the claimant had been one of two or more causes of the accident. Relying upon McCracken v Smith [2015] EWCA Civ 380, Stuart-Smith LJ stated that the doctrine of ex turpi causa would not apply in those circumstances.

The defence of volenti had been rejected at first instance. There was an inherent difficulty in the Ministry maintaining, on the one hand, that the height of the balustrade was not dangerous and, on the other, that the claimant knowingly accepted the very risk created by its inadequate height.

The broader lesson is that illegality, volenti and contributory negligence are not interchangeable labels for unattractive behaviour.

Illegality requires proper identification of the unlawful conduct and analysis of its relationship with the claim.

Volenti requires a genuine and voluntary acceptance of the relevant legal risk, not merely knowledge that an activity carries some danger.

Contributory negligence requires proof that the claimant failed to take reasonable care for his own safety and thereby contributed to the damage.

Over-pleading each defence may distract from the realistic issue, which is often a conventional assessment of the claimant’s own care and causative contribution.

Learning points for claimant solicitors

1. Establish the claimant’s status early

In employer-provided accommodation, obtain the employment contract, service requirements, accommodation rules and evidence showing whether residence was compulsory. The fact that the claimant was required to live in the barracks was central to the Court’s rejection of the temporary trespasser argument.

2. Identify the physical danger separately from the claimant’s conduct

Do not allow allegations of intoxication, horseplay or other misconduct to obscure the condition of the premises. Identify precisely what made the feature dangerous, who was foreseeably exposed to it and what reasonable precaution would have reduced the risk.

3. Obtain the complete risk-assessment history

Seek all pre-accident inspections and risk assessments, together with post-accident reports, learning documents, near-miss investigations and estate-wide reviews. A later document may reveal the nature of a pre-existing risk and what a competent earlier assessment should have identified.

4. Investigate changes of use

Establish the building’s original use, later use, refurbishment history and the characteristics of the present users. A structure built for one purpose may require reassessment when used for another.

5. Do not treat non-retrospectivity as determinative

Even where modern Building Regulations were not formally triggered, they may provide evidence of professional knowledge, accepted safety practice and the reasonableness of precautions.

6. Preserve compelling photographic evidence

Obtain photographs showing the guarding in context and, where appropriate, its relationship to the claimant or a suitable scale. Measurements alone may not adequately communicate the relationship between the height of the guarding, the claimant’s centre of mass and the risk of toppling.

7. Instruct the appropriate expert disciplines

A building surveyor may address construction, standards and condition. A suitably qualified engineer or falls expert may be required to address the mechanics of the fall and counterfactual causation. Do not assume that a single expert is necessarily equipped to cover both areas.

8. Frame expert questions around alternative mechanisms

Ask what would probably have happened with reasonably safe guarding under each factual mechanism supported by the evidence. This protects the case if the court rejects the claimant’s primary account but accepts an alternative version.

9. Address brain injury and memory expressly

Where the claimant’s recall is impaired, obtain appropriate neurological or neuropsychological evidence if required. Explain why reliability may be affected without suggesting dishonesty. Construct the case so that it does not depend entirely upon the claimant’s memory.

10. Treat early medical histories with care

Obtain the complete records and identify who gave each history, when it was taken and the claimant’s clinical condition. Do not conceal inconsistencies, but place them in their proper medical and evidential context.

11. Secure key witnesses promptly

If the defendant serves hearsay statements from identifiable witnesses, seek contact details and attendance without delay. Where appropriate, obtain permission to call them and consider witness summonses. Keep a clear record of any failure to provide information necessary to secure their attendance.

12. Require precision in any plea of contributory negligence

Ask the defendant to identify the precise act or omission relied upon, the danger which the claimant should have appreciated, the reasonable alternative course said to have been available, and the alleged causative effect.

Learning points for defendant solicitors

1. Test the danger directly

Do not rely solely upon the age of the building or compliance with historic standards. Ask whether the feature was reasonably safe for the building’s actual use and present users.

2. Avoid an exclusively technical Building Regulations defence

Non-retrospectivity may answer an allegation of direct regulatory breach. It does not, without more, answer a claim in negligence or under the Occupiers’ Liability Act 1957.

3. Preserve the complete documentary record

Secure all inspection, maintenance, accident, learning and risk-assessment materials. Internal documents which acknowledge danger or recommend remedial work may be more influential than the formal litigation position.

4. Consider carefully whether hearsay witnesses should be called

If the defendant serves witness statements as hearsay, it should decide whether the witnesses can and should attend. A party may face obvious forensic difficulty if it introduces untested evidence, fails to facilitate cross-examination and later attacks the reliability of its own material.

5. Do not build an appeal on a rejected factual premise

Every expert concession and legal submission must be linked to the findings actually made. A concession based upon a deliberate, forceful push did not assist where the judge found that no such push occurred.

6. Define any alleged change of visitor status precisely

If it is contended that the claimant became a trespasser, identify the relevant premises, the act said to exceed the permission, the point at which status changed and the legal basis upon which it changed. A duty which disappears for a few seconds and then reappears is unlikely to provide a workable analysis. It does not flicker on and off.

7. Plead illegality with restraint

Identify the offence, evidence and legal consequence. A breach of workplace rules, misconduct, criminality and contributory negligence are different matters and should not be conflated.

8. Obtain evidence directed to contributory negligence

Evidence should address the duration of the conduct, the parties’ position, any warnings, the claimant’s opportunity to disengage, his appreciation of the hazard and the causative effect of his actions. Moral criticism alone will not establish a deduction.

9. Consider whether a separate falls expert is required

If the claimant relies upon expert evidence on fall mechanics, the defendant should decide whether that evidence requires a specialist response. A decision not to serve the defendant’s own expert evidence may leave the claimant’s opinion substantially uncontroverted.

10. Put allegations fairly and expressly

Any allegation that the claimant is fabricating, exploiting convenient memory gaps or giving dishonest evidence must be put squarely in cross-examination. It should not emerge for the first time in closing submissions or on appeal.

A practical framework for stair and balustrade cases

The following sequence may assist practitioners on both sides.

Define the claimant’s legal status and permitted use of the premises.
Identify the physical feature alleged to create the danger.
Obtain accurate measurements, plans and good-quality photographs.
Establish the date of construction, original use, changes of use and relevant refurbishment.
Identify applicable regulations and standards, distinguishing legally binding requirements from evidence of contemporary safe practice.
Consider the characteristics of foreseeable users, including height, age, impairment, fatigue, distraction and foreseeable social behaviour.
Examine the adequacy of inspections and risk assessments before the accident.
Assess both likelihood and severity, rather than relying only upon the absence of previous accidents.
Identify the precaution which ought reasonably to have been taken, together with its cost, difficulty and likely effectiveness.
Reconstruct the mechanism from the whole body of evidence, distinguishing credibility from reliability.
Test causation against each realistic factual mechanism.
Analyse visitor status, volenti, illegality and contributory negligence separately rather than collapsing them into a general allegation of misconduct.

Conclusion

ZZZ is an important and factually rich authority. Its significance is not confined to military accommodation or alcohol-fuelled horseplay.

It confirms that the Occupiers’ Liability Act 1957 is concerned with the real relationship between an occupier and a visitor, not an artificial legal status which changes from second to second according to momentary conduct.

It confirms that older premises must be assessed in the light of their present use, their present users and current professional knowledge, even where modern regulations are not retrospectively binding.

It demonstrates that causation depends upon the factual mechanism found by the court, rather than the most dramatic hypothetical proposition put to an expert.

It shows that contributory negligence requires proper evidence and sufficiently precise findings, particularly where the relevant events were brief and the claimant’s memory has been impaired by brain injury.

The claimant’s success did not depend upon establishing that the Ministry was required to prevent young soldiers from drinking or engaging in every form of horseplay. It depended upon the more conventional proposition that an employer and occupier, providing compulsory accommodation to young soldiers, should have identified and remedied an obvious low balustrade which exposed occupants to the risk of a catastrophic fall.

Above all, the case is a reminder that a low balustrade is not merely part of the background to a fall. It may be the very feature which converts an ordinary stumble, a loss of balance or a brief moment of horseplay into a catastrophic brain injury.

The law does not demand a risk-free environment. It does require occupiers and employers to recognise an obvious, grave and readily remediable danger before somebody falls.

Please find a PDF version of the article here.

About the author

Pankaj Madan is a Barrister specialising in catastrophic injury with a specialist interest in brain injury. He acts for both Claimants and Defendants. He is the author of leading texts on Subtle Brain Injury and Chronic Pain and a recent book on Catastrophic Brain Injury. He practises from Exchange Chambers and 12 King’s Bench Walk. He is President-Elect of the Royal Society of Medicine, Pain Section Council. He is Legal 500 PI Junior of the year 2025.

Disclaimer and Exclusion of Liability

This article is provided for general information purposes only. It does not constitute legal advice and should not be relied upon as such. No responsibility or liability is accepted for any loss or damage arising from reliance on the contents of this article. Readers are strongly advised to seek appropriate, legal advice before taking any action or making any decisions based on the information provided herein. The author and publisher expressly exclude all liability for actions taken or not taken in reliance on this article.

Back to News