Thirty-two groups of four naive Ss and one accomplice participated in this 2 × 2 × 2 experiment designed to test conditions facilitating imitation of a competent model played by the accomplice. It was assumed that the psychological distance between the competent model and the Ss would be lessened so that imitation would increase: 1. if the model was friendly rather than hostile, 2. if the situation was non-frustrating rather than frustrating, and 3. if aggression was permitted rather than forbidden. Imitation was measured by the Ss' reproduction of the model's two mannerisms, and was found to be significantly influenced in the predicted way by the three variables. However, the freedom to aggress did not lessen psychological distance. Alternative explanations are provided for these results.