This research attempted to demonstrate some of the conditions that would influence the performance of generalized imitative behaviors in young children. Two experiments were conducted. The results of Exp. I indicated that generalized imitative behaviors can be very durable; only one of three subjects was influenced by a variety of reinforcement‐like procedures. Control over the behavior of all three subjects was obtained when a setting event involving the presence or absence of the experimenter was systematically varied. A second test of this variable was carried out in Exp. II. Results showed moderate to strong control over non‐reinforced imitations in four preschool child