Research on social facilitation over the 12 years since 196S is reviewed. It is concluded that the drive-theory analysis proposed by Zajonc in 1965 still provides the best overall theoretical framework for explaining social facilitation, but that Cottrell's elaboration, which emphasizes learned drives as the motivational basis of the phenomenon, appears justified. The main tenet of the drivetheory approach, that the presence of conspecific organisms is arousing, has received additional support from studies not based on Zajone's Hullian assumptions. The secondary motive state associated with social facilitation is probably aversive in nature and is describable in terms such as learned fear of failure, anxiety, or anticipatory frustration. Alternative explanations for social facilitation based on current cognitive views of behavior may ultimately shed light on important mediating processes but as yet do not possess the economy of constructs offered by the drive-theory approach.The study of social facilitation is as old as experimental social psychology itself. In the first social psychological laboratory investigation, Triplett (1898) found that speed on a simple motor task was greater among members of coacting pairs than among subjects performing alone. In the decades that followed, several investigators continued to report effects on performance of not only coactors but also passive audiences. The effects, however, were not always facilitative; often social settings produced performance decrements, thereby creating a paradox that remained essentially unsolved until the mid-1960s. Nonetheless, the collection of data on the problem proceeded at a brisk pace, peaking in the decade from 1925 to 1935. The early studies have been reviewed elsewhere (Cottrell, 1968(Cottrell, , 1972Dashiell, 1935) and will not be discussed here.Interest in social facilitation eventually waned. By 1954, Kelly and Thibaut could conclude that "these phenomena [i.e., audience and coaction effects], which were once thought to be basic to the study of social Requests for reprints should be sent to