Among 12,720 patients subjected to coronary angiography because of clinical suspicion of coronary heart disease, 2,234 (78% men and 22% women) were free from major modifiable coronary risk factors. They did not report alcohol consumption, use of antiplatelet or lipid lowering agents, oral contraceptives or sex hormones; there was no history of bilateral oophorectomy, smoking, diabetes mellitus, hypertension, or obesity; the ratio of total to high-density lipoprotein cholesterol was < or = 4.5. We examined lesions causing > or = 50% stenosis to total obliteration of the 3 great coronary arteries or of their major branches, and classified these patients as having single-, double-, or triple-vessel disease, or normal angiograms (no luminal irregularities). We related the extent of the coronary angiographic involvement to age and sex. From the fourth to the eighth decade of life: a) prevalence of normal angiography was significantly higher in women; b) percentage of cases with single-vessel disease was similar in both sexes and gradually reduced with aging; c) prevalence of double-vessel disease in women did not vary significantly, although a slight decline was seen at older ages; in men figures at the fifth and sixth decades were significantly greater than in women; d) percentages by decade of triple-vessel disease in males were 24, 34, 41, 49 and 57%, respectively; corresponding values in women were 11, 13, 15, 27 and 44% (p < 0.01 at each decade of age); with advances of age the sex gap in triple-vessel disease narrowed, but did not disappear. Thus, atherosclerosis tended with age to involve more vessels rather than more subjects in both genders, indicating that in the absence of modifiable coronary risk factors coronary vessels remained clearly susceptible to the influence of age, both in men and women. In these patients the feminine gender exerted a substantial protection that persisted for the entire span of life.