The principal–agent problem (often abbreviated agency problem) refers to the conflict in interests and priorities that arises when one person or entity (the "agent") takes actions on behalf of another person or entity (the "principal"). The problem worsens when there is a greater discrepancy of interests and information between the principal and agent, as well as when the principal lacks the means to punish the agent. The deviation of the agent's actions from the principal's interest is called "agency cost".

Common examples of this relationship include corporate management (agent) and shareholders (principal), elected officials (agent) and citizens (principal), or brokers (agent) and markets (buyers and sellers, principals). In all these cases, the principal has to be concerned with whether the agent is acting in the best interest of the principal. Principal-agent models typically either examine moral hazard (hidden actions) or adverse selection (hidden information).

The principal–agent problem typically arises where the two parties have different interests and asymmetric information (the agent having more information), such that the principal cannot directly ensure that the agent is always acting in the principal's best interest, particularly when activities that are useful to the principal are costly to the agent, and where elements of what the agent does are costly for the principal to observe.

The agency problem can be intensified when an agent acts on behalf of multiple principals (see multiple principal problem). When multiple principals have to agree on the agent's objectives, they face a collective action problem in governance, as individual principals may lobby the agent or otherwise act in their individual interests rather than in the collective interest of all principals. The multiple principal problem is particularly serious in the public sector.

Various mechanisms may be used to align the interests of the agent with those of the principal. In employment, employers (principal) may use piece rates/commissions, profit sharing, efficiency wages, performance measurement (including financial statements), the agent posting a bond, or the threat of termination of employment to align worker interests with their own.

The principal's interests are expected to be pursued by the agent; however, when the interests of the agent and principal differ, a dilemma arises. The agent possesses resources such as time, information, and expertise that the principal lacks. At the same time, the principal does not have control over the agent's ability to act in the agent's own best interests. In this situation, the theory posits that the agent's activities are diverted from following the principal's interests and drive the agent to maximize the agent's interests instead.

The principal and agent theory emerged in the 1970s from the combined disciplines of economics and institutional theory. There is some contention as to who originated the theory, with theorists Stephen Ross and Barry Mitnick both claiming authorship. Ross is said to have originally described the dilemma in terms of a person choosing a flavor of ice-cream for someone whose tastes they do not know (Ibid). The most cited reference to the theory, however, comes from Michael C. Jensen and William Meckling. The theory has come to extend well beyond economics or institutional studies to all contexts of information asymmetry, uncertainty and risk.

In the context of law, principals do not know enough about whether (or to what extent) a contract has been satisfied, and they end up with agency costs. The solution to this information problem—closely related to the moral hazard problem—is to ensure the provision of appropriate incentives so agents act in the way principals wish.^{[citation needed]}