There are several competing theories of proximate cause.
The most common test of proximate cause is foreseeability. It determines if the harm resulting from an action was reasonably able to be predicted. The test is used in most cases only in respect to the type of harm. It is foreseeable that throwing a baseball at someone could cause them a blunt-force injury. But proximate cause is still met if a thrown baseball misses the target and knocks a heavy object off a shelf behind them, which causes a blunt-force injury.
This is also known as the "extraordinary in hindsight” rule.
Direct causation is a minority test, which addresses only the metaphysical concept of causation. The main thrust of direct causation is that there are no intervening causes between an act and the resulting harm. An intervening cause has several requirements: it must 1) be independent of the original act, 2) be a voluntary human act or an abnormal natural event, and 3) occur in time between the original act and the harm.
Direct causation is the only theory that addresses only causation, and does not take into account the culpability of the original actor.
The plaintiff must demonstrate that the defendant's action increased the risk that the particular harm suffered by the plaintiff would occur. If the action were repeated, the likelihood of the harm would correspondingly increase. This is also called foreseeable risk.
The Harm Within the Risk test determines if the harm was within the class of risks, and if the victim was within the class of persons forseeably harmed. It is the strictest test of causation, made famous by Benjamin Cardozo in the Palsgraf case.
The first element of the test, if the harm was an instance of the type of harm that made the actor negligent, is met if the person could be expected to be hurt by the act.
A good illustration of this is that a pedestrian is within the class of people risked by driving on the sidewalk, whereas another driver who is distracted by someone driving on the sidewalk and thus crashes into a telephone pole is not.
The second prong of the test is if the type of harm was in the class of expected harms. Giving a loaded gun to a small child gives rise to the expectation that someone will be shot. If the child drops the gun and it breaks someone's foot, the HWR test will fail, as a broken foot was not the type of harm anticipated.
The HWR test is not used much anymore, and when it is used, just looks at the class of people hurt, not at the type of harm. The main issue with this test is that it addresses culpability, not actual causation.
The doctrine of proximate cause is notoriously confusing. The doctrine is phrased in the language of causation, but in most of the cases in which causation is contested, there is not much real dispute that the defendant but-for caused the plaintiff's injury. The doctrine is actually used by judges to limit the scope of the defendant's liability to a subset of the total class of potential plaintiffs who suffered some harm from the defendant's actions. For an understanding of the broader view of causation which proximate cause circumscribes, see Butterfly effect.
For example, in the two famous Kinsman Transit cases from the 2nd Circuit (exercising admiralty jurisdiction over a New York incident), it was clear that mooring a boat improperly could lead to the risk of a boat drifting away and crashing into another boat, and that both boats could crash into a bridge, which collapsed and blocked the river, and in turn, the wreckage could flood the land adjacent to the river, as well as prevent any traffic from traversing the river until it had been cleared. But under proximate cause, the property owners adjacent to the river could sue (Kinsman I), but not the owners of the boats or the cargoes which could not move until the river was reopened (Kinsman II).
Therefore, in 2001, the American Law Institute proposed in a draft of the Restatement (Third), Torts: Liability for Physical Harm (Basic Principles) that proximate cause should be replaced with scope of liability.
The term "proximate cause" is also used by historians, in the sense of a specific event or incident setting off a war, revolution etc. which had deeper roots and causes. An often cited example is the assassination of Archduke Franz Ferdinand of Austria at Sarajevo, the event which directly set off the First World War - but it is generally assumed by historians that the war was actually caused by the sharpening rivalries between the European powers and the setting up of rival military alliances, and that had the assassination not taken place, some other incident would have still set off the war. The distinction between an immediate proximate cause and the deeper and more fundamental causes has become a staple of historical research.