Pain intensity was observed before and after the subcutaneous (s.c.) administration of ketamine hydrochloride (80 micrograms/kg/s.c.) or saline (0.9% NaCl given by the same route) in 17 migraine (M) sufferers as an acute treatment of their M attacks. The same parameter was observed in another group of 17 M-sufferers complaining of very severe and frequent M attacks; these subjects were completely refractory to the prophylactic treatments currently used in M. In this second group, ketamine 80 micrograms/kg/s.c./three times a day) or saline was randomly assigned in a short (3-week) chronic treatment. A randomized, double-blind, cross-over study design was used both when testing ketamine as an acute administration for relieving M attack and when comparing its effect to that of the placebo in a chronic treatment for preventing M-pain. Ketamine, but not placebo, produced a marked relief of pain both as an acute treatment and as a prophylactic therapy. Mild specific side-effects were observed in the majority of the patients treated with ketamine. Moreover in the placebo group, the majority of these patients also complained of side-effects. The present results support the hypothesis that N-methyl-D-aspartic acid (NMDA) receptors play an important role in the mechanisms of M.