In a double-blind placebo-controlled trial, buflomedil was shown to cause a significant increase both in median claudication provoking time from 63 sec (range: 24-136 sec) to 124 sec (range: 53-261 sec) (p less than 0.01), and in maximum walking distance (MWD) from 169m (range: 157-308 m) to 293 m (range: 107-429 m) (p less than 0.01). The MWD after three months' buflomedil treatment was also significantly (p = 0.05) prolonged when compared with the MWD in the appropriate placebo group. In contrast, treatment with the placebo caused no significant change in these indices. Subjective improvement was observed in 12 out of 14 patients on buflomedil, whilst it occurred in only 6 out of 14 patients on the placebo (p less than 0.05). The clinical improvement was not associated with an increase in the ankle pressure index or a reduction in platelet aggregation and thromboxane A2 release.