This study sought to show that motivated biases can be shown for positive, as well as negative, behaviors. In the first study college students (N=151) estimated how often they and their fellow students engaged in various positive and negative health-related behaviors. Results indicated all negative behaviors yielded the predicted pattern: subjects believed they engaged in the behavior less often than did their average peer. The more important finding was that the bias observed for positive behaviors was substantially weaker than that observed for negative behaviors. In the second study college students (N=138) generated lists of either positive or negative health behaviors. In a subsequent questionnaire subjects rated how often they and their same-age, same-sex peers engaged in the behaviors they had listed. Of the 253 negative behaviors listed in which self and peer estimates differed, 77% yielded the predicted bias: subjects thought others committed these behaviors more often. By contrast, a significantly lower 50% of the 204 cases where self and peer estimates differed showed subjects estimating their own positive behaviors as more frequent than those of their peers. Once again, self-peer biases were stronger for negative behaviors, and there was not even evidence of bias for positive behaviors. (ABL)