Ten years ago Davison formulated the third-person effect hypothesis, a novel approach to the study of public opinion. Davison proposed that individuals typically assume that mass communications exert a stronger impact on others than the self, and he derived some interesting ideas from this notion. Over the past decade, a number of studies have tested predictions derived from Davison's formulation. This paper reviews and synthesizes research on the third-person effect. A systematic review of third-person effect studies indicates that there is abundant support for the notion that individuals assume that communications exert a stronger influence on others than on the self. However, the third-person effect does not emerge in all circumstances and for all people. The effect appears to be particularly likely to emerge when the message contains recommendations that are not perceived to be personally beneficial, when individuals perceive that the issue is personally important, and when they perceive that the source harbors a negative bias. Considerably less is known about the processes that underlie the third-person effect. This paper proposes several explanations for the effect, and it suggests some directions for future research in this area.