In a prospective study of about 16 000 men, serum samples were collected and stored. Vitamin-A (retinol) levels were later measured in the stored samples from the 86 men who were subsequently notified as having developed cancer and in the stored samples from 172 controls who did not develop cancer. Low retinol levels were associated with an increased risk of cancer. The association was independent of age, smoking habits, and serum-cholesterol level aand was greatest for men who developed lung cancer (mean retinol level 187 i.u./dl compared with 229 i.u./dl for the controls, p < 0.005). The risk of cancer at any site for men with retinol levels in the lowest quintile was 2.2 times greater than the risk for men with levels in the highest quintile (p < 0.025). These results suggest that measures taken to increase serum-retinol levels in man may lead to a reduction in cancer risk.