Treatment-emergent adverse events (AEs) were monitored during an 8-week, double-blind, placebo-controlled trial of risperidone (0.5-3.5 mg/day) in 101 children and adolescents with a lifetime diagnosis of autistic disorder. In addition, 37 placebo nonresponders received open-label risperidone for another 8 weeks. Of all the risperidone responders (n=65), 63 entered an open extension of another 16 weeks (6 months total risperidone exposure), and 32 of them were rerandomized to either continued risperidone therapy (n=16) or gradual replacement with placebo (n=16) over 8 weeks. We collected the following measures of safety and tolerability: (1) laboratory blood assessments (CBC with differential, electrolytes, and liver function tests) and urinalyses, (2) vital signs, (3) Side Effects Review of AEs thought to be associated with risperidone, (4) sleep records, (5) Simpson Angus Neurological Rating Scale (SARS), (6) Abnormal Involuntary Movement Scale (AIMS), and (7) height and weight. No clinically significant changes were found on the lab tests. During the 8-week acute trial, the most common AEs on the Side Effects Review, scored as moderate or higher, were as follows (placebo and risperidone, respectively): Somnolence (12% and 37%), enuresis (29% and 33%), excessive appetite (10% and 33%), rhinitis (8% and 16%), difficulty waking (8% and 12%), and constipation (12% and 10%). "Difficulty falling asleep" and anxiety actually favored the risperidone condition at statistically significant levels. The same AEs tended to recur through 6 months of treatment, although often at reduced levels. Using Centers for Disease Control (CDC) standardized scores, both weight and body mass index (BMI) increased with risperidone during the acute trial (0.5 and 0.6 SDs, respectively, for risperidone; 0.0 and 0.1 SDs, respectively, for placebo) and into open-label extension (0.19 and 0.16 SDs, respectively), although the amount of gain decelerated with time. Extrapyramidal symptoms, as assessed by the SARS, were no more common for drug than placebo, although drooling was reported more often in the risperidone group. There were no differences between groups on the AIMS. Two subjects had seizures (one taking placebo), but these were considered unrelated to active drug. Most AEs were mild to moderate and failed to interfere with therapeutic changes; there were no unanticipated AEs. The side effects of most concern were somnolence and weight gain.