ABSTRACT The present study examines the indirect "spillover" effects of methylphenidate on the behavior of the classmates of drug-treated children in public school settings, and it measures the extent to which medication normalizes hyperactive and noncompliant-aggressive behaviors. Eleven aggressive hyperactive boys, who met DSM-III diagnostic criteria for attention deficit disorder, received placebo and methylphenidate in a double-blind crossover design. Medication effects were assessed by conducting observations of the drug-treated boy and his peers in the classroom and lunchroom settings. In most classrooms, a spillover effect did not occur for most of the behaviors assessed, but there was a decrease in the rate of nonphysical aggression exhibited by peers as a function of the hyperactive child's methylphenidate dose. Peers were generally less aggressive when the hyperactive boys were receiving methylphenidate than when they received placebo. In the lunchroom, in contrast, peers appeared slightly more noncompliant and aggressive when the aggressive hyperactive boys were receiving medication compared with placebo, but these differences did not attain statistical significance. Methylphenidate effectively normalized the negativistic behaviors of the aggressive hyperactive boys in the classroom. In some cases, the frequency of occurrence of certain behaviors dropped significantly below the level of the hyperactive child's peers. This "supranormalization" may reflect behavioral toxicity of the psychostimulant.