Proximate cause is the legal rule that limits a defendant’s responsibility to harms that were a foreseeable type of risk from their conduct, even when their actions set a longer chain of events in motion. It is the point at which the law stops following consequences and says: this far, and no further. A careless act might ripple outward for blocks, but courts hold the actor accountable only for the outcomes closely and predictably tied to the risk they created.
What Proximate Cause Actually Decides
Proximate cause is a policy tool, not a physics equation. Factual causation can trace an injury back through a long sequence of events, but at some point the connection becomes too remote to fairly pin on the person who started it. Proximate cause draws that line.
Many legal professionals call this concept “legal cause” instead, which is a more honest label. Courts are making a judgment about fairness, not tracking billiard balls. The question is whether the link between conduct and harm is tight enough that it makes sense to force the defendant to pay.
The Restatement (Third) of Torts has moved away from the term “proximate cause” entirely, replacing it with “scope of liability.” Under that framework, the central question is whether the harm that occurred was the same general type of harm that made the defendant’s conduct risky in the first place. A growing number of courts have adopted this language, though the underlying analysis is largely the same. Whether a court calls it proximate cause or scope of liability, it is asking whether this particular harm falls within the risks that made the behavior negligent.
The Foreseeability Test
The dominant tool for analyzing proximate cause is foreseeability. The question is whether a reasonable person in the defendant’s position could have anticipated that their conduct would create a risk of the type of injury that actually occurred. The key word is “type.” Courts do not require that the defendant foresaw the exact chain of events or the precise mechanism of injury, only that the general category of harm was predictable.
If a contractor leaves an uncovered hole in a sidewalk, it is foreseeable that someone will fall in and get hurt. The contractor does not need to have predicted that a specific 72-year-old woman would trip at 3 p.m. on a Tuesday and fracture her hip. The general type of harm, a pedestrian falling, is what matters.
Palsgraf and the Zone of Danger
The 1928 case of Palsgraf v. Long Island Railroad Co. remains the most influential American decision on foreseeability. A railroad employee helped a passenger board a moving train, and in the process the passenger dropped a package of fireworks. The explosion knocked over a heavy scale at the far end of the platform, injuring a woman named Helen Palsgraf. She sued the railroad.
Chief Judge Benjamin Cardozo, writing for the majority, held that Palsgraf could not recover because she was not within the foreseeable zone of danger created by the employee’s conduct. The employee’s action might have been careless toward the passenger carrying fireworks, but “nothing in the situation gave notice that the falling package had in it the potency of peril to persons thus removed.” The risk that defined the duty was the risk to the package holder, not to a bystander standing far away. No foreseeable risk to Palsgraf meant no duty owed to her, and without a duty, her negligence claim failed.1New York State Law Reporting Bureau. Palsgraf v Long Island Railroad Co
The Andrews Dissent
Judge Andrews wrote a famous dissent that still shapes how some courts think about proximate cause. He argued that everyone owes a general duty to the world to refrain from unreasonably dangerous conduct, and that once a negligent act occurs, the defendant should be liable to anyone actually injured by it, not just those within a foreseeable zone of danger. Under Andrews’ view, the question was whether the harm was a direct and natural consequence of the act, considering factors like whether the chain of events was continuous, whether the result was too remote in time and space, and whether the defendant’s conduct was a substantial factor in producing it.1New York State Law Reporting Bureau. Palsgraf v Long Island Railroad Co
Cardozo’s foreseeability approach won the day and became the dominant standard in American tort law. Andrews’ “direct consequences” reasoning survives in some jurisdictions and occasionally resurfaces when courts feel the foreseeability test produces unjust results.
Proximate Cause Is Not the Same as Cause in Fact
A plaintiff must prove two separate types of causation. Cause in fact asks whether the defendant’s conduct physically produced the injury. Proximate cause then asks whether the law should attach liability to that conduct given the circumstances. Both must be established.
The standard test for cause in fact is the but-for test: would the injury have occurred but for the defendant’s act? A driver who runs a red light and hits a pedestrian passes this test easily. The pedestrian would not have been hit but for the driver running the light.
The but-for test breaks down when two independent forces each would have been sufficient on their own to cause the harm. The classic example involves two fires, one set negligently and one caused by lightning, that merge and burn down a house. Neither fire is a but-for cause, because the house would have burned anyway from the other. Applying the test strictly would let both defendants escape. Courts handle this with the substantial factor test, which asks whether the defendant’s conduct was a significant contributor to the harm. If so, the defendant is liable even though another force would have produced the same result independently.
Clearing both tests only establishes that the defendant physically caused the injury. Proximate cause is the separate question of whether the law considers that physical connection close enough to impose liability.
Intervening and Superseding Causes
Other events sometimes occur between the defendant’s negligent act and the plaintiff’s injury. An intervening cause is any event that happens in that gap and contributes to the harm. An intervening cause does not automatically let the defendant off the hook. What matters is whether the intervening event was foreseeable.
Foreseeable intervening causes leave the defendant’s liability intact. Medical errors during treatment of an accident injury are the textbook example. If a driver causes a crash and the victim’s surgeon makes a mistake during a follow-up operation, the original driver typically remains liable for the additional harm. Medical complications from treating injuries are a known risk of getting injured, so they do not break the causal chain.
A superseding cause is different. It is an intervening event so extraordinary and unforeseeable that it replaces the defendant’s negligence as the legal cause of the harm. When a court labels something a superseding cause, the original defendant’s liability ends. The event must be both independent of the original negligence and beyond what anyone could reasonably have anticipated. A sudden earthquake that collapses a building on an already-injured plaintiff might qualify. A random criminal attack by a stranger with no connection to the original incident might qualify too, depending on the circumstances.
The line between a foreseeable intervening cause and a superseding one is where many cases are won or lost. Courts look at how unusual the intervening event was, whether it was a natural response to the situation the defendant created, and how closely connected it was to the original risk.
The Rescue Doctrine
Rescuers get special treatment. When a defendant’s negligence puts someone in danger, it is foreseeable that a bystander will try to help. The rescue doctrine holds that the defendant is liable for injuries the rescuer suffers during a reasonable rescue attempt. The classic formulation, recognized by Cardozo himself in an earlier opinion, is that “danger invites rescue.”
The doctrine prevents a defendant from arguing that the rescuer’s decision to intervene broke the chain of causation. Courts treat the rescuer as a foreseeable plaintiff so long as the rescue attempt was not reckless or wildly disproportionate to the danger. A bystander who runs into a burning car to pull someone out is protected. A bystander who sprints across a freeway to retrieve a stranger’s hat probably is not.
The Eggshell Skull Rule
Foreseeability determines whether the defendant is liable at all. Once that threshold is crossed, the eggshell skull rule governs how much they owe. This doctrine says a defendant must take the victim as they find them. If the plaintiff has an unusually fragile skull, a bleeding disorder, or a pre-existing back injury, the defendant pays for the full extent of the harm, even if a healthier person would have walked away with a bruise.
This produces an important distinction. The type of harm must be foreseeable, but the severity does not. A defendant who rear-ends another car can foresee that the other driver might suffer a neck injury. If that driver happens to have a degenerative spinal condition that turns a minor whiplash into paralysis, the defendant is still on the hook for the paralysis. The argument that most people would not have been hurt that badly is legally irrelevant. Between an innocent plaintiff with a pre-existing vulnerability and a negligent defendant, the defendant bears the risk of unexpectedly severe outcomes.
How Proximate Cause Fits Into a Negligence Case
Proximate cause is one of four elements every negligence plaintiff must prove: a duty of care owed by the defendant, a breach of that duty, causation (both factual and proximate), and actual damages. The plaintiff carries the burden of proving each element by a preponderance of the evidence, meaning it is more likely true than not. That is a lower bar than the “beyond a reasonable doubt” standard used in criminal cases, but it still requires more than speculation.
For straightforward cases, the causal link may be obvious enough that no special proof is needed. A driver rear-ends you at a stoplight and your neck hurts afterward. The connection between the impact and the injury is within any juror’s common experience. When the causal chain gets longer or more technical, expert testimony becomes essential. Medical malpractice cases almost always require a physician to explain how the provider’s specific error caused the patient’s specific harm, because jurors lack the medical knowledge to evaluate that connection on their own.
Expert witnesses do not just testify that harm occurred. They must connect the defendant’s conduct to the harm with reasonable medical or scientific probability. A mere possibility that the defendant’s actions caused the injury is not enough. The expert must explain why the defendant’s breach, rather than other potential causes, was the likely explanation for the plaintiff’s condition. Without this testimony in technical cases, courts routinely grant summary judgment for the defendant, ending the case before it ever reaches a jury.
The final determination of proximate cause is usually a question for the jury. The judge decides whether enough evidence exists for a reasonable jury to find proximate cause; if so, the jury weighs that evidence against the defendant’s counterarguments. All the legal theory about foreseeability, intervening causes, and scope of liability ultimately gets filtered through the people in the jury box, who decide whether it makes sense to hold this defendant responsible for this particular harm.
A Note on Your Own Negligence
Proving proximate cause does not guarantee full compensation. If the plaintiff’s own carelessness contributed to the injury, most jurisdictions reduce the damage award proportionally. A plaintiff found 20% at fault for a $100,000 injury recovers $80,000. In many states, crossing a fault threshold (typically 50% or 51%) bars recovery entirely, and a handful of jurisdictions still apply the older contributory negligence rule, which blocks recovery if the plaintiff was even slightly at fault. A case can clear every proximate cause hurdle and still produce a smaller recovery than the raw damages suggest.