Twenty-four patients with Raynaud's phenomenon were treated with griseofulvin and a placebo in a double-blind cross-over clinical trial. The results were assessed subjectively and plotted sequentially. The number of patients able to distinguish the active from the placebo treatment was statistically significant (P=0·05). There were indications that the response was dose dependent. Further studies are needed to establish whether griseofulvin offers useful clinical improvement in this condition.