In an impression explanation of counterdefensive attributions, the tendency of subjects is to accept primary responsibility for negative outcomes and to deny credit for positive outcomes. Counterdefensive attributions represent attempts at maximizing the esteem in which observers hold the actor. To test this hypothesis, a 2 x 5 simulation was carried out of Ross, Bierbrauer and Polly's (1974) experiment in which professional teachers gave "moderately counterdefensive" attributions for what they believed to be their successful or unsuccessful teaching performance. Simulation observers (N=100) learned that a particular teacher had either succeeded or failed and that he either did not explain or had offered one of four different explanations for the performance outcome (a highly or moderately defensive or counterdefensive explanation). The results strongly supported the impression management analysis. In both failure and success conditions, observers evaluated the teacher most positively--on ability and personality traits--when he gave moderately counterdefensive attributions. Defensive attributions consistently elicited the most negative observer evaluations. (Author)