We compared two prompting techniques that are commonly used to teach individuals with autism. In the "most-to-least" (MTL) prompting condition, the therapist initially delivered the most intrusive prompt necessary to achieve a correct response. Prompts were gradually faded across subsequent trials, while errors resulted in the provision of increasingly more intrusive prompts. Prompt fading occurred in the same manner for the "no-no-prompt" (NNP) condition; however, the therapist presented the initial instruction up to two times before delivering a prompt. Four children with autism participated. Rate of skill mastery, frequency of errors, and maintenance of skills under the two prompting conditions were compared via combined multielement and multiple baseline designs. Although the NNP method resulted in faster skill acquisition, MTL prompting was associated with fewer errors for all participants and better performance during 1- and 2-week maintenance probes for 3 of the 4 participants. (Contains 1 table and 2 figures.)