The purpose of the study was to address the hypothesis that elevated body iron increases the risk of cancer occurrence and death, and to determine the dose response. Subjects were 3,287 men and 5,269 women participating in the first National Health and Nutrition Examination Survey who had a transferrin saturation determination at enrollment (1971-1975), who remained alive and cancer-free for at least 4 years, and who were followed to 1988 for cancer outcome. Among 379 men who developed cancer over the study period, the mean transferrin saturation at enrollment was 32.1% whereas among 2,908 who remained cancer-free it was 30.7%; the difference for mortality was 32.3% among 233 deaths vs. 30.8% among 3,054 men not dying of cancer. The mean differences among women were not significant. The mean differences in TIBC and serum iron among men were consistent with the findings for transferrin saturation, and all 3 differences were stable over time when examined by years since blood test. Men and women were divided into 5 groups on the basis of baseline transferrin saturation: 0 to 30%, 30-40%, 40-50%, 50-60%, and 60% and higher. Nineteen percent of men had a baseline transferrin saturation above 40% (the last 3 groups), whereas only 10 percent of women had transferrin saturation above 40%. For men and women combined, risk of cancer occurrence in each group relative to the first was 1.0, 0.95, 1.16, 1.38 and 1.81; for mortality the relative risks were 1.0, 0.96, 1.22, 1.29 and 1.73. There is evidence, in this cohort, of elevated cancer risk in those with moderately elevated iron level. This pattern was seen in women as well as in men.