No pharmacological treatment protocol has proven generally useful for all patients who stutter. Various medications, behavior therapy, relaxation, suggestion, and social-based therapies have been used. For this drug treatment study, two groups of adult stutterers were followed in an 8-week open label protocol. All subjects had in the past received speech therapy; none had been treated previously with medication for stuttering. The first group (N=12) received a maximum dose of 800 mg of carbamazepine; the second group (N=8) received a maximum dose of 400 mg of carbamazepine. Each patient served as his or her own control. A series of systematic speech tests was given weekly to determine the variability of fluency for each subject.A statistically significant change occurred for a number of “expectancy to stutter” characteristics. Subjects felt that they stuttered less often while taking carbamazepine. Subjective effects began before medication and continued after patients discontinued the medication. Struggle characteristics also subjectively decreased. However, no objective improvement was found. No change was found in percentage of words stuttered, reading improvement, or improvement in spontaneous speech rate. Interrater reliability showed a correlation of .996. Three carbamazepine serum level therapeutic windows were inspected with negative results. Interestingly, naive listener ratings did show a statistically significant improvement on carbamazepine versus placebo. Future anecdotal reports of pharmacological improvement of stuttering should be subjected to rigorous objective testing before general acceptance