Capillary microscopy is an easy, noninvasive procedure to examine in vivo the dermis capillaries of the nailfold. It has demonstrated clinical use in the etiologic study of Raynaud's phenomenon. We consider a "vascular activity pattern" to be present at capillary microscopy when one of the following associations is observed: tortuous capillary vessels plus ramifications with or without loss of a moderate amount of capillaries. Capillary tortuousity alone is not considered pathologic. It is well known that bleomycin (BLEO) can occasionally induce vascular-associated diseases. To examine the vascular damage produced by BLEO, we performed capillary microscopic studies on 40 patients with neoplasia, 21 of whom had received BLEO during the previous year. The maximum accumulated doses ranged from 15 to 379 U. The other 19 patients had advanced neoplasia, and 10 of them had received antitumoral combinations that did not contain BLEO. No one had clinical signs attributable to vascular toxicity. Twenty-six patients had no pathologic patterns (11 received BLEO and 15 did not). In the group of 14 patients with activity patterns, 10 had received BLEO and 4 had not (P = 0.035). Ten of 11 patients treated with BLEO who had normal capillary microscopic studies received total doses of less than 100 U. We conclude that capillary microscopy may demonstrate the vascular damage induced by BLEO even in asymptomatic patients.