We study strategic interactions that may be affected by belief-dependent, conformist preferences. Specifically, we propose that beliefs about the behavior of individuals in the same role (i.e., beliefs about “peer behavior”) directly affect a player’s utility. In examining conformism we propose an experimental design that verifies the presence of the relevant causality direction. Our data reveal “opportunistically conformist” behavior, as subjects are more likely to follow the purported majority if doing so implies an increase in expected material payoff. We provide a general framework that accounts for such a pattern.