Children with autism and related disabilities frequently fail to develop echoic repertoires. Among the ways in which treatment approaches vary is the extent to which automatic reinforcement is utilized. The present experiment was designed to test the efficacy of a procedure that incorporates automatic reinforcement and socially mediated reinforcement in the development of an echoic repertoire. The implementation of this treatment package resulted in an increase in the vocal play and echoic behavior for two participants, each of whom had very limited verbal repertoires. The study was conducted as a multiple probe across sounds design. Vocal play for Participant 1 increased from baseline rates as low as 0.2 per min to 4.5 per min during treatment, and her percent correct echoic behavior increased from 0% to at least 90% on both sounds. Participant 2 showed rapid gains in echoic control on one sound before requiring dismissal from the study. Interobserver agreement on the occurrence of target vocals equaled 100%. The results of this study have implications for which procedures to use in the establishment of echoic, echoic mand, and mand repertoires. (Contains 2 figures and 1 footnote.)