The purpose of this investigation was to determine whether instructional behavior learned during a series of simulated teaching experiences using peers as students would transfer to a setting in which "real" pupils were students. Subjects were 44 undergraduate students in the secondary education program at the University of Texas at Austin. Each teacher presented a series of lessons with peers serving as students for the simulated teaching and one lesson with sixth, seventh, and eighth grade pupils as students. Four lessons were rated and coded--the first lesson, the lesson immediately preceding that using "real" students, the real pupils lesson, and the lesson immediately following that using "real" pupils. Ratings were made of the four teaching dimensions--determining readiness, clarifying objectives, motivating, and evaluating. Results indicated some behavior change during the period of simulated teaching with peers and provided some evidence that instructional behavior acquired during simulated teaching with peers will transfer to a simulated setting using actual pupils. It cannot be concluded that these results would be obtained in other settings, such as the classrooms in which these teachers will eventually be placed. The results suggest that when peers are used as students in simulated teaching, attempts should be made to occasionally use actual pupils in the simulated teaching experience. (MBM)