What judgmental standards guide the evaluation of social conduct? Four studies were designed to show that conduct evaluations are made egocentrically by holding others up to standards of behavior that correspond to people's estimates of how they would resolve ethical dilemmas. Consistent with this assumption, subjects in Experiment I who evaluated targets who resolved ethical decision dilemmas in the same manner as themselves blamed the target less than subjects who endorsed the opposite resolution. Experiments 2, 3, and 4 were designed to rule out the possibility that this relative evaluation effect arose from people's perceptions that their own behavioral preferences were in some sense evaluatively superior to the alternative, and hence that people were actually imposing a single standard (perceptions of right and wrong) rather than different, egocentric standards (i.e., their own behavioral choices), on others. These studies found that those who selected "nonmoral" alternatives (e.g., cheating on a test) diminished blame for targets who did likewise, even though they judged the moral alternatives to be more ideally correct, evaluatively "better," and would rather have their children select these alternatives. The implications of these results for legal applications of the "reasonably prudent person" standard, and for self-reference effects in social judgment, were elaborated.