In a controlled trial of amitriptyline hydrochloride in migraine prophylaxis, 100 patients received placebo for a four-week baseline period and then were randomized in double-blind fashion to therapy with amitriptyline (47 subjects) or placebo (53 subjects) for another four to eight weeks. Subjects received up to four 25-mg tablets of amitriptyline hydrochloride or identical placebo per day. Comparing the first and second four-week periods for each patient, the conditions of 55.3% of amitriptyline subjects as opposed to 34.0% of placebo subjects were greater than or equal to 50% improved and the difference between amitriptyline and placebo response rates was significant (P less than .05). Nondepressed subjects with severe migraine and depressed subjects with less severe migraine responded best to amitriptyline, whereas depressed subjects with severe migraine had little headache relief. Amitritryline is an effective antimigraine agent and the antimigraine effect seems relatively independent of antidepressant activity.