Forty-six patients with hand dystonia, considered disabling despite optimal pharmacologic therapy, were injected in the forearm musculature with botulinum A toxin. Thirty of these patients were followed long enough to provide adequate data for analysis of 86 treatment sessions. There was a 63% female preponderance, with an average age at initial evaluation of 46 years and symptom duration of 7.9 years. Average baseline severity of dystonia was rated as 3.5 on a severity rating scale (0-4 rating; 4 = maximum severity). The average peak effect response for all injections (79 into wrist flexors and 29 into wrist extensors) was 2.2 for dystonia and 3.0 for pain (0-4 rating; 0 = no response, 4 = maximum benefit). The latency from injection to onset of effect averaged 5.6 days. Total response duration averaged 9.3 weeks and maximum improvement was 7.5 weeks. Only local complications occurred and consisted primarily of hand weakness (25 patients, 44 sessions). The results show that botulinum toxin injections effectively control hand dystonia in instances where other forms of therapy have failed.