OBJECTIVE To test the hypothesis that intramuscular (IM) botulinum toxin type A (BTX) reduces excessive muscle tone in a dose-dependent manner in the elbow, wrist, and fingers of patients who experience spasticity after a stroke. DESIGN Randomized, double-blind, placebo-controlled, multicenter, 24-week trial. SETTING Six academic and 13 private US outpatient medical centers. PARTICIPANTS Ninety-one patients with a mean age of 60 years (range, 30-79 y). Mean time elapsed from ischemic or hemorrhagic stroke to study enrollment was 25.8 months (range, 0.9-226.9 mo). INTERVENTIONS Up to 2 treatments of placebo, or 90, 180, or 360U of BTX. Concurrent splinting and physical therapy protocols were permitted, but no changes were allowed during the study. MAIN OUTCOME MEASURES Wrist, elbow, and finger flexor tone assessed by the Modified Ashworth Scale, physician and patient global assessments, pain, FIM instrument, and Medical Outcomes Study 36-Item Short-Form Health Survey (SF-36). RESULTS Muscle tone decreased more with injections of BTX than with placebo in the wrist flexors at weeks 1, 2, 3, 6, and 9 (P< or =.026); in the elbow flexors at weeks 1, 2, 3, 4, 5, and 9 (P< or =.033); and in the finger flexors at weeks 1 and 3 (P< or =.031). A dose-dependent response was generally observed in tone reduction but not in pain, FIM, or SF-36 measures. CONCLUSIONS IM BTX reduced muscle tone in a dose-dependent manner in the elbow, wrist, and fingers of patients who experience spasticity after a stroke but did not appear to affect global quality of life or disability.