BACKGROUND AND PURPOSE: The rationale for our study was to examine the prophylactic benefit of two doses of amitriptyline over a 6-month observational period in patients with migraine. We aimed at evaluating whether 50 mg of amitryptiline extended release was more effective than 25 mg in reducing the number of migraine days. METHODS: Primary outcome measure was the reduction of migraine days in time course (i.e., 3 and 6 months after patient enrolment). As secondary analyses, predictors of treatment response were evaluated. Treatment response was defined as reduction of > or =30% and > or =50% in migraine headache days in time course. RESULTS: The intent-to-treat population comprised 132 patients (female 96; male 36) with migraine. Median migraine days per month were reduced from 7 days (range: 6-15) at baseline, to 6 days (range: 4-12; P < 0.001) at 3 months, and to 6 days (range: 3-12; P < 0.001) at 6 months, respectively. However, no statistically significant difference in the number of migraine days was seen between the two treatment groups at 3 and 6 months. As a result of secondary analyses, the number of migraine days per month at baseline was the only independent predictor of response to amitriptyline treatment (for both definitions of treatment response, i.e., response rate > or =30% and response rate > or =50%) at 6 months. CONCLUSIONS: The prophylactic effect of amitriptyline seen in our study was rather weak and did not differ between the two treatment groups. The results of this 6-month, prospective, open-label clinical observation are therefore not encouraging.