Psychostimulants and behavior therapy have been postulated to be effective in treating attention-deficit hyperactivity disorder (ADHD) by compensating for a pathologically elevated reward threshold, but no studies have compared reinforcement to psychostimulants in maintaining task performance. The separate and combined effects of methylphenidate (MPH, 0.6 mg/kg) and a behavioral intervention (reward plus response cost) were assessed on a continuous performance test (CPT, a measure of sustained attention) modified to deliver auditory feedback contingent upon the subject's responses. Each of 22 children (6-10 years old) with ADHD were tested under four treatment conditions: placebo + feedback, placebo + behavioral contingencies, MPH + feedback, and MPH + contingencies. CPT performance, indexed by d' (ability to discriminate between target and false targets), was significantly better with MPH than with placebo, showing reduced deterioration over time. Contingency treatment improved mean d' compared to placebo + feedback but, in contrast, had no effect on the slope of performance deterioration. Addition of contingencies to MPH did not yield further improvement. The results indicate that MPH improved sustained attention on a laboratory task (and reduced task-irrelevant and other disinhibited behaviors), whereas behavioral contingencies did not. These findings suggest that, although both interventions improved stimulus discrimination processes, only MPH enhanced processes that mediate the regulation of effort over time. In addition, the disjunction between the effects of reward and of MPH provides evidence that psychostimulant effects on attention are only partially explained by the stimulation of brain centers associated with reward.