The interactive effects of implicit normative and explicit situational consensus information were examined regarding the processes of causal attribution and evaluation. Stimulus items were single sentence descriptions of antisocial and prosocial behaviors representing the extremes of high and low normative consensus in each behavior category, as determined from estimates given by a pretest group of "the proportion of people in society who would be likely to perform the behavior described." High and low levels of situational consensus information were created by inserting standard phrases into the sentences indicating that the actor either had many or few co-actors in his immediate context. Subjects (N=151) chose the behavior's "main cause" from three options and then evaluated the actor using a nine-point bipolar scale. It was found that internal attribution frequencies were affected by situational but not by normative consensus, while external attribution frequencies were affected by both kinds of consensus information. Attribution responses also varied as a function of behavior category. Evaluations were affected by normative but not by situational consensus information. Behavior category influenced extremity of evaluation at the low normative consensus level only. Also, internal attributions were generally followed by more extreme evaluations than were external attributions. (Author)