We enrolled 55 patients in a double-blind, placebo-controlled, parallel design study of the effectiveness of botulinum toxin (Botox) injections for the treatment of spasmodic torticollis. Patients received a standard series of injections, either placebo or Botox. We determined the sites of injection and dose per muscle by the nature of head deviation. Compared with placebo, Botox produced statistically significant improvement in the severity of torticollis, disability, pain, and degree of head turning. There were no serious side effects. During the double-blind phase, 61% of patients injected with Botox improved; 74% of patients subsequently improved during a later open phase at a higher dose of Botox. Direction of head turning, severity of torticollis, and presence or absence of jerky movements did not significantly influence the response rate. We conclude that Botox is a valuable treatment for spasmodic torticollis.