This study examines 1) the effects of teacher contingent and non-contingent (random) praise and/or attention on the classroom behavior of economically disadvantaged adolescents and 2) the effects of contingent praise on the non-target members of the classroom, and explores the length of time necessary to demonstrate significant changes in behavior. The subjects were 150 eighth-grade high school students in Knoxville, Tennessee, and six teachers with several years experience. Three behavioral categories were established: appropriate, relevant, and inappropriate. Six graduate students served as observers and for each class three target and three non-target students were observed daily. The results show that contingent teacher praise and/or attention was effective in controlling behavior. Teacher praise or attention to relevant behavior in combination with ignoring disruptive behaviors increased relevant behavior and concomitantly reduced inappropriate behaviors. Non-contingent praise increased the percentage of relevant behavior for target subjects but not to the extent of contingent praise. A closer examination of the data reveals that there were individual subject differences, and generalizations about the rapidity or amount of changes in classroom behavior occurring within a specified time period are tenuous. The implication of the study is that a teacher can modify and influence the behavior of his students if he can control his own behavior. (Author/MBM)