Repetitive behaviors are a core symptom domain in autism that has been linked to alterations in the serotonin system. While the selective serotonin-receptive inhibitor fluvoxamine has been shown to be effective in adults with autism, as yet no published placebo controlled trials with these agents document safety and efficacy in children with autism. This study examines the selective serotonin reuptake inhibitor liquid fluoxetine in the treatment of repetitive behaviors in childhood and adolescent autism spectrum disorders (ASDs). In total, 45 child or adolescent patients with ASD were randomized into two acute 8-week phases in a double-blind placebo-controlled crossover study of liquid fluoxetine. Study design included two randomized 8-week fluoxetine and placebo phases separated by a 4-week washout phase. Outcome measures included measures of repetitive behaviors and global improvement. Low-dose liquid fluoxetine (mean final dose: 9.9+/-4.35 mg/day) was superior to placebo in the treatment of repetitive behaviors by CY-BOCS compulsion scale. The effect size was in the moderate to large range, and the doses used were low. Liquid fluoxetine was only slightly, and not significantly, superior to placebo on CGI autism score partially due to a phase order effect. However, fluoxetine was marginally superior to placebo on a composite measure of global effectiveness. Liquid fluoxetine did not significantly differ from placebo on treatment emergent side effects. Liquid fluoxetine in low doses is more effective than placebo in the treatment of repetitive behaviors in childhood autism. Limitations include small sample size and the crossover design of the study. Further replication and long-term maintenance trials are needed.