Three children with selective mutism were treated with a packaged procedure that incorporated self-modeling, mystery motivators, self-reinforcement, stimulus fading, spacing, and, in one case, antidepressant medication. All three children evidenced a complete cessation of selective mutism and maintained their treatment gains at follow-up. An argument was presented for the use of a series of single-case, baseline-intervention (AB) designs appropriate for this type of school-related, lour incidence behavior. Specific conditions were met that allowed for inferences to be drawn that approached the quality of those typically obtained from experimental designs. (C) 1998 Society for the Study of School Psychology. Published by Elsevier Science Ltd.