Abductive reasoning (also called abduction, abductive inference, or retroduction) is a form of logical inference that seeks the simplest and most likely conclusion from a set of observations. It was formulated and advanced by American philosopher and logician Charles Sanders Peirce beginning in the latter half of the 19th century.

Abductive reasoning, unlike deductive reasoning, yields a plausible conclusion but does not definitively verify it. Abductive conclusions do not eliminate uncertainty or doubt, which is expressed in terms such as "best available" or "most likely". While inductive reasoning draws general conclusions that apply to many situations, abductive conclusions are confined to the particular observations in question.

In the 1990s, as computing power grew, the fields of law, computer science, and artificial intelligence research spurred renewed interest in the subject of abduction. Diagnostic expert systems frequently employ abduction.

Deductive reasoning allows deriving ${\displaystyle b}$ from ${\displaystyle a}$ only where ${\displaystyle b}$ is a formal logical consequence of ${\displaystyle a}$. In other words, deduction derives the consequences of the assumed. Given the truth of the assumptions, a valid deduction guarantees the truth of the conclusion. For example, given that "all men are mortal" (${\displaystyle a_{1}}$) and "Socrates is a man" (${\displaystyle a_{2}}$), it follows that "Socrates is mortal" (${\displaystyle b}$).

Inductive reasoning is the process of inferring some general principle ${\displaystyle b}$ from a body of knowledge ${\displaystyle a}$, where ${\displaystyle b}$ does not necessarily follow from ${\displaystyle a}$. ${\displaystyle a}$ might give us very good reason to accept ${\displaystyle b}$ but does not ensure ${\displaystyle b}$. For example, if it is given that 95% percent of the elephants are gray, and Louise is an elephant, one can induce that Louise is gray. Still, this is not necessarily the case: 5 percent of the time this conclusion will be wrong.

However, an inference being derived from statistical data is not sufficient to classify it as inductive. For example, if all swans that a person has observed so far are white, they may instead abduce the possibility that all swans are white. They have good reason to believe the conclusion from the premise because it is the best explanation for their observations, and the truth of the conclusion is still not guaranteed. (Indeed, it turns out that some swans are black.)

Abductive reasoning allows inferring ${\displaystyle a}$ as an explanation of ${\displaystyle b}$. As a result of this inference, abduction allows the precondition ${\displaystyle a}$ to be abducted from the consequence ${\displaystyle b}$. Deductive reasoning and abductive reasoning differ in which end, left or right, of the proposition "${\displaystyle a}$ entails ${\displaystyle b}$" serves as the conclusion. For example, a couple leaving their house in the morning and seeing that their lawn is wet might abduce that it rained while they were asleep. This serves as a hypothesis that "best explains" their observation. Given the many possible explanations for the lawn getting wet, their abduction does not establish certainty that it rained overnight, but it is still useful and can serve to orient them in their surroundings. Despite many possible explanations for any physical process we observe, we tend to abduce a single explanation (or a few) for this process, in the expectation that we can better orient ourselves in our surroundings and disregard some possibilities. Properly used, abductive reasoning can be a useful source of priors in Bayesian statistics.

One can understand abductive reasoning as inference to the best explanation, although the terms abduction and inference to the best explanation are not always used equivalently.