The "But For" Test in Irish Law: Factual Causation in Personal Injury Claims

Definition Capsule: The But-For Test at a Glance

Doctrine

The but-for test of factual causation

Origin

English common law, formally articulated in Barnett v Chelsea & Kensington Hospital Management Committee [1969] 1 QB 428

Status in Ireland

The primary test for factual causation in tort

Leading Irish authority

Quinn v Mid Western Health Board [2005] IESC 19

Standard of proof

Balance of probabilities (the plaintiff's burden)

Typical application

Medical negligence, road traffic, employer's liability, public liability — every tort claim where causation is in dispute

When it fails

Where the harm would have occurred regardless of the breach (see Barnett; see Crumlish v HSE [2024] IECA 244)

Contents

Quick Definition: The But-For Test in One Paragraph

The but-for test asks a single counterfactual question: but for the defendant's wrongful act or omission, would the plaintiff still have suffered the same injury? If the harm would have happened anyway — independently of the breach — the defendant did not cause it in law, and the claim fails on causation. The test operates at the stage of factual causation, after duty of care and breach of duty have been established. It must be proven by the plaintiff on the balance of probabilities, the standard civil burden in Irish tort proceedings.

Origin and Adoption in Irish Law

The but-for test traces its modern formulation to the English Queen's Bench decision in Barnett v Chelsea & Kensington Hospital Management Committee [1969] 1 QB 428, decided by Nield J. Three night-watchmen attended the casualty department of the hospital on the morning of 1 January 1966, complaining of vomiting after drinking tea that had been laced with arsenic. They spoke to a nurse, who relayed their complaints by telephone to the casualty officer, Dr Banerjee. Dr Banerjee did not examine the men in person; he advised the nurse over the phone that they should go home and consult their own doctors. Mr Barnett, one of the three, died of arsenic poisoning some five hours later. The hospital's failure to examine the men was a breach of duty. But Nield J held that, even with prompt examination and admission, there was little or no chance that the antidote could have been administered in time to prevent Mr Barnett's death. The court found the hospital not liable: the breach had not caused the death, because the death would have occurred regardless.

That counterfactual logic — strip out the negligence and ask whether the outcome changes — has become the orthodox starting point for factual causation across the common law world, including in Irish jurisprudence. In Ireland, the Supreme Court treated it as the controlling test in Quinn v Mid Western Health Board [2005] IESC 19, where Kearns J (Denham J and Geoghegan J concurring) declined to displace the but-for approach in favour of broader alternative tests imported from English mesothelioma jurisprudence.

Factual Causation and Legal Causation: A Two-Stage Inquiry

Causation in Irish negligence is not a single question but a two-stage inquiry. The but-for test is the first stage — factual causation — and asks whether the breach made a difference in the actual sequence of events. A claim that fails on factual causation goes no further. A claim that passes still has to clear the second stage — legal causation, also called remoteness — which asks whether the kind of harm that occurred was reasonably foreseeable as a consequence of the breach. The orthodox foreseeability test traces to Overseas Tankship (UK) Ltd v Morts Dock & Engineering Co ("The Wagon Mound (No 1)") [1961] AC 388, and is applied in Irish courts as a separate filter on liability.

The practical effect is that a defendant whose breach satisfies the but-for test may still escape liability if the harm was of a kind no reasonable person would have anticipated. Conversely, a foreseeable kind of harm that would have happened anyway — Mr Barnett's death from arsenic poisoning being the textbook illustration — produces no liability, because factual causation fails at the first stage. The two-stage structure is doctrinal architecture, not pleading nicety: each stage carries its own evidence, its own burden, and its own appellate treatment.

How the But-For Test Works in Practice

In practice, the but-for test operates as the gateway to liability in any tort case where factual causation is contested. Once duty of care is established and breach is proven (under the Dunne principles in medical cases, or the reasonable-person standard generally), the plaintiff must then show that the breach made a difference to the outcome. The test is conducted as a counterfactual reconstruction: the court considers what would have happened in the hypothetical world where the defendant exercised reasonable care, and compares that outcome to the actual injury suffered.

Three structural features shape how the test plays out in Irish courtrooms:

1. The burden remains on the plaintiff. The Supreme Court in Quinn was explicit that any approach which had the effect of reversing the onus of proof would be a change of such magnitude as to require a full court — or perhaps even legislation — before it could be made. The plaintiff must tilt the scales decisively in favour of the proposition that the breach changed the outcome.

2. The standard is the balance of probabilities, not certainty. The plaintiff need not prove that the harm would inevitably have been avoided but for the breach; only that, more likely than not (greater than 50%), the breach caused or materially contributed to the injury actually suffered. In practice, this turns on expert medical or engineering evidence about what the realistic counterfactual looks like.

3. The test is performed at trial, not at pleading. Causation is a question of fact, decided on evidence. The Court of Appeal's recent decision in Crumlish v HSE [2024] IECA 244 illustrates how a claim can pass every preliminary stage and still fail at "the first causation hurdle" once expert evidence is rigorously tested.

Worked Examples: Applying the But-For Test

The three illustrations below are composite scenarios drawn from common Irish fact-patterns. They show how the counterfactual analysis runs in practice — one passing the test, one failing it, and one in the contested middle ground where alternative causation theories arise. None is a real case; they are for educational illustration only.

Leading Irish Cases Applying the But-For Test

The cases below are the principal Irish authorities on the operation of the but-for test in personal injury and medical negligence. Each is decided by the Supreme Court or Court of Appeal and remains good law as at the date of last review.

When the But-For Test Does Not Apply

The but-for test works cleanly when there is one defendant, one breach, and one harm. It struggles in three distinct situations: where multiple causes are independently sufficient to produce the harm; where multiple causes cumulatively produce the harm and science cannot apportion their contributions; and where the breach has deprived the plaintiff of a chance rather than caused a defined outcome. Irish law's response to each situation is different, and the differences matter.

Multiple Sufficient Causes (Overdetermination)

The textbook overdetermination problem — two independent fires meet at the property, either of which alone would have destroyed it — exposes a logical gap in strict but-for reasoning. Applied literally, the test produces the perverse result that neither fire-setter caused the loss, because the property would have burned down anyway. Irish courts (in line with the wider common law) treat each defendant's contribution as causative in such cases by relaxing the strict counterfactual: where the breach was a sufficient cause, the analysis does not require it to be the necessary cause. The doctrinal route used to reach this result is usually material contribution to injury, not a true reformulation of but-for.

Material Contribution to Injury (the Bonnington Approach)

Where a single indivisible injury is caused by a combination of tortious and non-tortious exposures (or by exposures from multiple tortious sources), and science cannot say which exposure produced which portion of the harm, the plaintiff is permitted to establish causation by proving that the defendant's breach materially contributed to the injury. The principle traces to the House of Lords decision in Bonnington Castings Ltd v Wardlaw [1956] AC 613, where Lord Reid framed the test in terms of a contribution that is more than de minimis. Bonnington-style material contribution does not displace the but-for test; it operates as a relaxation of the counterfactual where strict application would defeat a meritorious claim. Irish courts have applied the principle in industrial-disease and occupational-injury contexts.

Material Increase in Risk (the Fairchild Approach — Not Adopted)

The English position has gone further. In McGhee v National Coal Board [1973] 1 WLR 1 and most prominently in Fairchild v Glenhaven Funeral Services [2002] UKHL 22, the House of Lords held that where the defendant's breach materially increased the risk of an indivisible injury (mesothelioma), and science could not identify which exposure was actually causative, the plaintiff could recover against any defendant whose breach contributed to the risk. This is a different and more radical departure from but-for reasoning: the plaintiff is not required to prove the breach contributed to the injury, only to the risk of injury.

The Irish Supreme Court in Quinn declined to adopt the Fairchild approach. The court was offered the chance to apply the material-increase-in-risk theory to a clinical-causation problem and rejected it, holding that the orthodox principles relating to causation, "properly applied," were adequate to the task. This remains the controlling Irish position.

Recent Development: Tolan v Brindley Manor [2025] IEHC 327

A note of qualification has been sounded at High Court level. In Tolan v The Brindley Manor Federation of Nursing Homes Ltd [2025] IEHC 327, an interlocutory ruling reported through practitioner-publication channels (and confirmed by the case-listing of senior counsel acting for the defendant nursing home), Barr J declined to dismiss a clinical negligence claim that relied on material-contribution arguments at the strike-out stage. The judgment is treated by Irish practitioners as a signal that the courts may engage more flexibly with material-contribution theories in the right factual setting, though it does not displace the controlling Supreme Court authority in Quinn.

Citation note: The full text of the Tolan judgment was not located in the standard primary repositories (courts.ie or BAILII) at the date of last review. Verification rests on the published case-list of Rory White SC at the Bar of Ireland Law Library and on contemporaneous practitioner commentary; an updated direct link will be added when the judgment is published in primary sources.

Loss of a Chance

Where the negligence has deprived the plaintiff of a measurable chance of a better outcome, rather than causing a defined injury, the orthodox but-for test cannot be applied directly — the counterfactual is itself probabilistic. Philp v Ryan [2004] IESC 105 supports limited recovery on this basis. The position differs from England and Wales, where the House of Lords in Gregg v Scott [2005] UKHL 2 rejected loss-of-chance recovery in clinical negligence where the patient's initial survival probability had not exceeded 50%. The relationship between Irish loss-of-chance recovery and the orthodox but-for test is not fully settled and continues to attract practitioner commentary, most recently in the Law Society Gazette analysis of November 2024.

The But-For Test Outside Medical Negligence

Although clinical negligence cases produce the most contested causation arguments, the but-for test governs every Irish tort. The mechanics shift with the factual setting, but the core counterfactual — what would have happened had the defendant exercised reasonable care? — is constant. Three settings illustrate the range.

Road Traffic Cases

In a road traffic action, the counterfactual is usually constructed by an accident-reconstruction engineer. If the alleged breach is excessive speed, the question becomes whether the impact would have occurred at all (or would have produced the same injuries) at the speed limit. If the alleged breach is a failure to look properly before manoeuvring, the question becomes whether a reasonable lookout would have detected the hazard in time to react. The expert evidence reconstructs stopping distances, perception–reaction times, and impact dynamics; the trial judge applies the but-for test to that reconstruction.

Two practical points recur. First, where the plaintiff has not worn a seatbelt or helmet, the defendant typically argues the failure to wear protective equipment is a separate but-for question — would the injuries have occurred, or been as severe, with the equipment worn? — and pleads contributory negligence under section 34 of the Civil Liability Act 1961 in the alternative. Second, in cases involving multiple vehicles, each defendant's contribution is analysed separately at the but-for stage before the Civil Liability Act's apportionment machinery is engaged.

Employer's Liability and Workplace Cases

In a workplace accident claim, the counterfactual is built around the employer's non-delegable duty to provide a safe system of work. If the breach is a failure to provide manual-handling training, the question is whether the injury would have occurred had the training been provided — which in turn requires evidence about whether trained employees doing the same task suffer the same kind of injury. If the breach is a failure to maintain machinery, the question is whether the failure caused the equipment fault that produced the injury. Expert evidence on workplace ergonomics, machinery maintenance, or industry standards frames the counterfactual.

The but-for test in this setting is overlaid by the statutory framework of the Safety, Health and Welfare at Work Act 2005 and its underlying regulations, which can establish breach more readily than the common-law test alone — but causation still has to be proved separately.

Occupier's Liability and Public Liability Cases

In a slip-or-trip claim, the counterfactual asks whether the injury would have occurred had the occupier exercised the duty owed to the relevant entrant category under the Occupiers' Liability Act 1995. If the breach is a failure to clean a wet floor, the question is whether reasonable cleaning frequency would have prevented the spillage being present at the moment of the fall. If the breach is a failure to install handrails, the question is whether handrails would have prevented the loss of balance. The Court of Appeal's slip-and-trip jurisprudence repeatedly emphasises that the plaintiff must establish the causation link, not simply the existence of the hazard.

The But-For Test in Medical Negligence

Causation is the most common reason otherwise-strong medical negligence claims fail in Ireland. In practice, medical negligence cases turn on causation more often than they turn on breach, because the State Claims Agency's standard defence strategy — particularly in delayed-diagnosis and birth-injury cases — frequently concedes (or at least does not strongly contest) breach, and instead concentrates its expert resources on the counterfactual. Practitioners refer to this gateway as "the first causation hurdle," picking up Noonan J's framing in Crumlish.

Three patterns recur:

Delayed-diagnosis cancer claims. The plaintiff must establish, on expert oncology evidence, that earlier diagnosis would have led to a materially better outcome. Crumlish shows how this can collapse where the claim depends on a contested mathematical model of tumour growth. The defence in such cases typically deploys what Irish practitioners colloquially call the Goldilocks problem: a slow tumour-doubling time means the cancer was not growing aggressively, so the delay made little clinical difference; a fast doubling time means the cancer was undetectable at the earlier date. The plaintiff is squeezed between the two — too slow and the delay does not matter, too fast and the tumour was not yet there to be found. The framework is set out in more detail in our practice-side guide on causation in medical negligence.

Birth-injury claims. The defendants typically accept some failure of monitoring or delayed delivery but argue, on neonatology and paediatric-neurology evidence, that the brain injury had already occurred before the breach point. Quinn is the controlling authority, and its facts remain the textbook illustration of why early-action arguments must be supported by causation evidence as well as breach evidence — a point developed further in our guide on birth injury negligence.

Failure-to-refer claims. The plaintiff must establish that referral, had it been made, would have produced an investigation that would have detected the condition at a stage where treatment was available and would have changed the outcome. The chain has multiple links, and each is vulnerable to the same but-for analysis. Our guide on failure-to-diagnose claims sets out how the chain is constructed in pleadings.

If I were convinced that defendants who have a substantial connection to the injury were escaping liability because plaintiffs cannot prove causation under currently applied principles, I would not hesitate to adopt [an alternative]. In my opinion, however, properly applied, the principles relating to causation are adequate to the task.

per Kearns J in Quinn v Mid Western Health Board [2005] IESC 19

But-For and Contributory Negligence: Section 34 of the Civil Liability Act 1961

The but-for test asks whether the defendant's breach caused the plaintiff's injury. A different question — whether the plaintiff's own conduct contributed to the injury — is governed by section 34 of the Civil Liability Act 1961. The two questions interact, but they are doctrinally distinct: the but-for test is about whether liability arises at all; section 34 is about how damages are apportioned once liability is established.

The mechanics work in this order. First, the court applies the but-for test to the defendant's breach. If causation fails at this stage, the case is over and section 34 is never reached. Second, if causation is established, the court considers whether the plaintiff's own act or omission also caused or contributed to the injury — applying the same counterfactual logic to the plaintiff's conduct. Where the plaintiff's contribution is established, section 34(1) requires the court to reduce damages by such amount as the court "thinks just and equitable having regard to the degrees of fault of the plaintiff and defendant."

The interaction matters most in three recurring scenarios. In seatbelt cases, the defendant argues that even if the collision was their fault on a strict but-for analysis, the severity of the injuries was caused by the plaintiff's failure to wear a belt — a separate but-for inquiry that founds a section 34 reduction (typically 15–25%). In intoxicated-passenger cases, the defendant argues that the plaintiff's decision to travel with an obviously impaired driver contributed to the injuries, with reductions commonly in the 25–40% range. In workplace cases, the defendant argues that the plaintiff's failure to wear provided protective equipment, or to follow a known safe-system-of-work step, contributed to the injury.

The Supreme Court confirmed in Hussey v Twomey [2009] IESC 1 that the test for the plaintiff's contribution is objective — what a reasonable person in the plaintiff's position would have done — not subjective to the particular plaintiff's perception. The doctrinal effect is that the but-for analysis runs twice in a contributory negligence case: once against the defendant's breach, and once against the plaintiff's own conduct.

Frequently Asked Questions

What does "but for" actually mean in a negligence claim?

It means: take the defendant's mistake out of the picture, and ask whether the plaintiff would still have suffered the same injury. If yes, the mistake didn't cause the injury in law and the claim fails on causation, even if the mistake itself was clearly negligent.

The test is applied after duty of care and breach of duty have been established. It is the third element of any negligence action and is decided by the trial judge on the medical, engineering or other expert evidence presented at trial. The standard is the balance of probabilities — more likely than not, not beyond reasonable doubt.

Practitioner note: In Irish medical negligence practice, the State Claims Agency's defence strategy is built around the causation question. Expect every weak link in the counterfactual to be probed by the defendant's experts.

Read more: See our practice-side guide on causation in medical negligence for how this plays out in real claims.

Is the but-for test the same in Ireland as in the UK?

The core test is identical — both jurisdictions trace it to Barnett v Chelsea & Kensington HMC [1969] 1 QB 428. The divergence is at the edges: English courts have developed broader alternatives (material contribution to risk in Fairchild; material contribution to injury in Bailey) that the Irish Supreme Court declined to follow in Quinn v Mid Western Health Board [2005] IESC 19.

Practically, this means Irish plaintiffs have less room than English plaintiffs to argue alternative causation theories where the strict but-for test cannot be satisfied. The position in Ireland on loss of chance, by contrast, is somewhat more generous than the post-Gregg v Scott position in England and Wales.

Practitioner note: Irish practitioners citing English causation authorities should be alert to which propositions are actually part of Irish law. Fairchild is sometimes invoked by plaintiffs but has not been adopted.

What is the leading Irish case on the but-for test?

Quinn v Mid Western Health Board [2005] IESC 19 is the modern leading authority. It confirms that the but-for test is the primary causation standard in Irish tort law, that the burden of proof rests at all times on the plaintiff, and that Irish courts will be cautious about adopting alternative causation tests imported from English jurisprudence.

The case concerned a catastrophic birth injury where the defendants conceded negligence but established on the evidence that the harm was caused by an acute event for which they were not responsible. The Supreme Court dismissed the appeal.

Practitioner note: Quinn is the case to cite in any contested Irish causation argument. Crumlish v HSE [2024] IECA 244 is the modern Court of Appeal application of the same principles.

When does the but-for test not apply?

The strict test is supplemented in three recognised situations: (i) where the defendant's breach materially contributed to the injury alongside other causes (the Bonnington approach, occasionally applied in Irish industrial-disease contexts); (ii) where the breach has deprived the plaintiff of a measurable chance of a better outcome (recognised in a limited form by Philp v Ryan [2004] IESC 105); and (iii) the English material-increase-in-risk approach in Fairchild, which has not been adopted in Ireland.

Outside these recognised situations, the orthodox but-for test controls and the plaintiff must satisfy it on the balance of probabilities.

Practitioner note: The boundary between the strict test and the "material contribution" gloss remains contested in Irish jurisprudence. The most accessible recent practitioner discussion is the Law Society Gazette analysis from November 2024.

Why do so many medical negligence claims fail on causation?

Because in clinical practice, breach and harm are often separated by an intervening medical pathway whose counterfactual is genuinely uncertain. A delayed-diagnosis cancer claim, for example, requires the plaintiff to prove that earlier diagnosis would have produced a materially better outcome — but cancer biology, treatment response, and patient-specific factors all introduce uncertainty into that counterfactual.

The State Claims Agency's defence strategy concentrates expert resources on this question, and the Irish courts (most recently in Crumlish) have shown they will dismiss claims where the causation evidence does not survive rigorous scrutiny, even where breach is otherwise plain.

Read more: Our guides on causation in medical negligence and expert medical reports set out the practical evidence framework.

Does the burden of proof on causation ever shift to the defendant?

Not under current Irish law. The Supreme Court in Quinn rejected an invitation to reverse the onus, holding that any change of that magnitude would require a full court or legislation. In contrast, the English courts have permitted limited reversals in mesothelioma cases under Fairchild and the Compensation Act 2006 (UK).

The plaintiff therefore carries the burden at all stages: in pleadings, in the exchange of expert reports, and at trial. A failure to assemble convincing causation evidence is, in practice, fatal to the claim.

Practitioner note: This is one of the clearer points of divergence between Irish and English causation law and should be flagged in any pleading that draws on English authorities.

What is the difference between factual causation and legal causation?

Factual causation is the but-for question: would the harm have happened anyway, without the defendant's breach? Legal causation — also called remoteness — is a separate filter that asks whether the kind of harm that occurred was reasonably foreseeable as a consequence of the breach. A claim has to clear both gates.

The distinction matters because each stage carries its own evidence and its own appellate treatment. A defendant can lose on factual causation (the breach made a difference) but win on legal causation (the kind of harm was so unforeseeable that the law will not impose liability for it). The orthodox foreseeability test traces to The Wagon Mound (No 1) [1961] AC 388 and is applied by Irish courts as the second stage of the inquiry.

Practitioner note: Pleadings should keep the two stages structurally separate. Conflating them is a common drafting error and complicates appellate review where one finding is challenged but not the other.

Does the but-for test apply outside medical negligence?

Yes. The but-for test governs every Irish tort, including road traffic, employer's liability, occupier's liability, and product liability. The mechanics differ — the counterfactual in an RTA case is built by an accident-reconstruction engineer, while in a workplace case it is built by an ergonomics or machinery expert — but the core question is the same: would the injury have occurred had the defendant exercised reasonable care?

The contested causation arguments concentrate in medical negligence because clinical pathways are biologically uncertain and lend themselves to plausible counterfactual disputes. In road traffic and workplace cases, the counterfactual is usually more tractable (impact dynamics, machinery faults), so causation is contested less often than in clinical claims.

Practitioner note: The but-for test in non-clinical cases is overlaid by statutory frameworks (the Safety, Health and Welfare at Work Act 2005; the Occupiers' Liability Act 1995) that affect breach but not causation. Causation must still be proven separately on common-law principles.

References

1. Barnett v Chelsea & Kensington Hospital Management Committee [1969] 1 QB 428; [1968] 2 WLR 422; [1968] 1 All ER 1068 — High Court of Justice, Queen's Bench Division (England and Wales), Nield J. Reported judgment; cited as the foundational articulation of the but-for test in modern common law.
2. Quinn (a minor) v Mid Western Health Board [2005] IESC 19, [2005] 4 IR 1; [2005] 2 ILRM 180 — Supreme Court of Ireland, Kearns J delivering judgment with Denham J and Geoghegan J concurring, 8 April 2005.
3. Philp v Ryan [2004] IESC 105 — Supreme Court of Ireland, Fennelly J delivering judgment, 16 December 2004.
4. Crumlish v Health Service Executive [2024] IECA 244 — Court of Appeal of Ireland, Noonan J delivering judgment with Power J and Binchy J, 15 October 2024.
5. Tolan v The Brindley Manor Federation of Nursing Homes Ltd [2025] IEHC 327 — High Court (Barr J); interlocutory ruling on motion to dismiss/stay. Full judgment text not located in primary online repositories at the date of last review; verified through the published case-list of Rory White SC at the Bar of Ireland Law Library and contemporaneous practitioner commentary.
6. Hussey v Twomey [2009] IESC 1 — Supreme Court of Ireland; controlling Irish authority on the objective test for contributory negligence under section 34 of the Civil Liability Act 1961.
7. Bonnington Castings Ltd v Wardlaw [1956] AC 613 — House of Lords (UK); foundational material-contribution-to-injury authority cited in Irish industrial-disease jurisprudence.
8. McGhee v National Coal Board [1973] 1 WLR 1 — House of Lords (UK); material increase in risk authority, considered but not adopted by the Irish Supreme Court in Quinn.
9. Fairchild v Glenhaven Funeral Services Ltd [2002] UKHL 22 — House of Lords; mesothelioma material-contribution-to-risk authority, declined in Ireland in Quinn.
10. Gregg v Scott [2005] UKHL 2 — House of Lords; English position on loss-of-chance recovery in clinical negligence.
11. Overseas Tankship (UK) Ltd v Morts Dock & Engineering Co ("The Wagon Mound (No 1)") [1961] AC 388 — Privy Council; the foreseeability test for legal causation (remoteness) applied as the second-stage filter in Irish negligence cases.
12. Kate Ahern, "Untangling the threads" — Law Society Gazette, November 2024. Practitioner analysis of material contribution doctrine in Irish causation law.
13. Civil Liability Act 1961, Act No. 41 of 1961 — Office of the Attorney General, irishstatutebook.ie. Section 34 governs apportionment for contributory negligence.
14. Occupiers' Liability Act 1995, Act No. 10 of 1995 — Office of the Attorney General, irishstatutebook.ie. Statutory duty framework for occupier and recreational-user/trespasser cases.
15. Safety, Health and Welfare at Work Act 2005, Act No. 10 of 2005 — Office of the Attorney General, irishstatutebook.ie. Statutory framework underlying employer's-liability breach analysis.