The purpose of this study was to test whether the effects of methylphenidate on rates of operant responding in hyperactive boys differed depending on the predrug baseline rates of responding. On the basis of the rate dependency hypothesis, dose-related decreases from high baseline rates and increases from low baseline rates would be predicted. A multiple FR-DRL (fixed ratio-differential reinforcement of low Rates) schedule was used to generate high and low rates of responding on a simple operant task, using nickels as reinforcers. Three dosage levels were used: placebo, 0.3 mg/kg, and 1.0 mg/kg of methylphenidate. Both experimenter and subject were blind to dosage level. Surprisingly, DRL response rates were not significantly affected by the drug, while FR rates increased linearly with increasing dose. Thus, the predicted between-schedule differences were not found. However, within the FR schedule, changes from baseline to postdrug rates within each active drug condition were consistent with the rate dependency hypothesis; i.e., subjects with lower baselines increased their response rates, while subjects with higher baselines decreased or did not change their rates. The opposite effect occurred following placebo.