ABSTRACT Clinical trials with a treatment duration of at least 3 months were reviewed to determine the effect of psychostimulants on the core symptoms of attention-deficit hyperactivity disorder (ADHD) and on its commonly associated features. Eighteen studies were identified: 17 were studies of methylphenidate, 1 was a study of dextroamphetamine, and none involved pemoline or slow-release stimulants. Eleven of these studies were randomized controlled trials, whereas seven employed quasi-experimental designs without randomization. The results of randomized controlled trials showed that psychostimulants provided greater benefit than did the nonrandomized trials, suggesting that the efficacy of extended treatment may have been underestimated because more seriously disturbed children were assigned to medication treatment than to control treatments in nonrandomized trials. Evidence from the more definitive randomized controlled trials indicates that stimulants are more effective in ameliorating the core behavioral symptoms of ADHD (restlessness, inattentiveness, impulsiveness) than placebos, nonpharmacological therapies, or no treatment-at least in 3-7-month trials. During extended psychostimulant treatment, few children become symptom-free, clinical effects may diminish with time, and improvement dissipates rapidly upon discontinuation of medication. There is minimal evidence that extended stimulant treatment improves cognitive deficits or associated problems such as conduct disturbance, low self-esteem, poor peer relationships, or academic underachievement. The belief that stimulants do not improve the long-term prognosis of children with ADHD may be based on weakly designed studies that focus on associated rather than core symptoms. A failure to assign patients of similar clinical severity to different treatment conditions may also have contributed to obscuring the efficacy of extended psychostimulant treatments.