Thirty patients (mean age: 10.38 years) affected by primary headache were selected for a double-blind cross-over clinical trial. The patients were randomized into 2 homogeneous groups of 15 and treated for 12 weeks with L-5-HTP (100 mg/day) and placebo as per the following design: placebo - L-5-HTP (group A) and L-5-HTP - placebo (group B). Evaluation was carried out every 3 weeks by the Migraine Index supplying a general assessment of the attacks, i.e. severity, duration and frequency. The decrease in mean score values was directly proportional to L-5-HTP treatment, and statistical significance (Wilcoxon's test) was observed only for L-5-HTP in both groups, from 0.05 to 0.01. Improvement, as evaluated by CGI on percentage distribution of the patients, was homogeneous in both groups.