Three microteaching formats were compared: microteaching with peers, microteaching with university freshmen, and microteaching with high school pupils. Thirty social studies methods students were assigned to one of the three formats for six experiences. Each experience included periods of instruction and practice. Assessments were made at the end of microteaching by a panel of graduate assistants using the Illinois Teacher Performance Appraisal Scale, and at the end of student teaching by pupils using the Illinois Teacher Evaluation Questionnaire. Significant differences at the completion of microteaching favored peer teaching, but microteaching with high school pupils was associated with superior performance at the end of student teaching. It appears easier to obtain desired training effects through peer teaching, but these effects do not seem to transfer to student teaching. (Author/RT)