Objective: In the United States approximately 750,000 children receive psychostimulants to treat attention-deficit hyperactivity disorder (ADHD); up to 25% may not respond. The purpose of this study was to evaluate the reports in the international literature concerning the efficacy of carbamazepine (CBZ) in children with ADHD features by means of meta-analysis.

Method: A review of the world literature located 29 reports that dealt with behavior problems, activity levels, and CBZ in children. Of these, only 10 reports provided sufficient or pertinent information for the meta-analysis.

Results: In all seven open studies, therapeutic responses were significant (ranging from p = .05 to .0001, two-tailed t test). Meta-analysis using weighted variables revealed a significant positive correlation (r = .88; p < .02) between duration of treatment and positive outcome. In three double-blind placebo-controlled studies, treatment effects for CBZ's superiority over placebo ranged from p = .07 to .0001. A meta-analysis of these three studies revealed that CBZ was significantly (p = .018) more effective than placebo at controlling target symptoms.

Conclusions: Despite the general lack of attention that CBZ has received for treating ADHD, there is preliminary evidence that CBZ may be an effective alternate treatment in children with features of ADHD.