Indirect effects occur when the impact of one species on another requires the presence of a third species. They can arise in two general ways: through linked chains of direct interactions, and when a species changes the interactions among species. Indirect effects have been uncovered largely by experimental studies that have monitored the response of many species and discovered ''unexpected results,'' although some studies have looked for specific indirect effects predicted from simple models. The characteristics of such approaches make it likely that the many indirect effects remain uncovered, but the application of techniques such as path analysis may reduce this problem. Deterministic theory indicates that indirect effects should often be important, although stochastic models need exploration. Simulation models indicate that some indirect effects may stabilize multi-species assemblages. Five simple types of indirect effects have been regularly demonstrated in nature: exploitative competition, trophic cascades, apparent competition, indirect mutualism, and interaction modifications. Detailed experimental investigations of natural communities have yielded complicated effects. Indirect effects have the potential to affect evolutionary patterns, but empirical examples are limited. Future directions in the study of indirect effects include developing techniques to estimate interaction strength in dynamic models, deriving more efficient approaches to detecting indirect effects, evaluating the effectiveness of approaches such as path analysis, and investigating mechanisms in which a species changes how other species interact.