Attacks of chronic paroxysmal hemicrania are prevented by the continuous administration of indomethacin. Sumatriptan, an agonist of 5-HT1-like receptors, has proven effective in the treatment of cluster headache attacks. There are clear clinical similarities between chronic paroxysmal hemicrania and cluster headache. A natural consequence of these considerations would be to establish whether chronic paroxysmal hemicrania also responds similarly to sumatriptan. Since hemicrania continua is another unilateral headache responsive to indomethacin, it would be meaningful to also include hemicrania continua in such a study. Sumatriptan, 6 mg subcutaneous, was tried in an open fashion in 7 patients (6 women and 1 man) with chronic paroxysmal hemicrania and 7 patients (5 women and 2 men) with hemicrania continua. In chronic paroxysmal hemicrania, the mean interval between the last three attacks prior to sumatriptan treatment (40 +/- 23 minutes) was not statistically different from the mean interval between the three attacks subsequent to sumatriptan treatment of an attack (32 +/- 20 minutes). In none of the patients did the mean duration of the "test attack" decrease as compared to the attacks antedating the test attack (25 +/- 11 minutes and 19 +/- 9 minutes, respectively) (P = 0.027, Wilcoxon). In 2 patients with chronic paroxysmal hemicrania, placebo (saline) administration did not lead to any change in the interval between attacks. There was a mild, but statistically significant reduction in visual analog scale values for headache intensity in hemicrania continua (P = 0.04, Wilcoxon). There was no clear, i.e., clinically meaningful, reduction in visual analog scale values in any particular patient with hemicrania continua. Taken together, these results seem to show that sumatriptan is of no benefit in chronic paroxysmal hemicrania, but may have a partial efficacy in hemicrania continua. However, the latter effect is clinically unimportant. This minor difference in regard to the clinical effect may, nevertheless, be of some interest pathogenetically, indicating minor differences between the two headaches. The lack of sumatriptan effect in chronic paroxysmal hemicrania clearly and markedly strengthens the nonalignment concept in regard to chronic paroxysmal hemicrania and cluster headache.