Because supplements of vitamins C and E had been associated with reduction of fecal mutagen levels, a double-blind randomized trial was designed to examine the effects of these vitamins on the rate of recurrence of colorectal polyps, presumed precursors for colorectal cancer. Two hundred patients believed to be free of polyps after removal of at least one colorectal polyp were randomized to receive a supplement of 400 mg each of ascorbic acid and alpha-tocopherol, or a placebo. Fifteen patients had to be excluded because a review of pathology indicated that their polyps were not adenomatous. A second colonoscopic examination was planned after 2 yr of supplementation. One hundred thirty-seven people (75% of eligible subjects) completed the study; polyps were observed in the second colonoscopy in 41.4% of 70 subjects on vitamin supplements and in 50.7% of 67 subjects on placebos. After adjustment for differences between groups in demographic and dietary factors before study entry, the relative risk of polyp occurrence was 0.86, with 95% confidence limits from 0.51 to 1.45, in an analysis of 129 subjects with complete information on demographic and dietary factors who had completed the trial. Of the 48 patients who had not completed the study, 7 had died, 33 had not returned to their physician for an examination, and 8 had had a follow-up colonoscopy or sigmoidoscopy. Inclusion of the three polyps found in these eight examinations led to an estimate of relative risk of 0.86 (95% confidence limits, 0.51 to 1.43). The findings of this investigation suggest that any reduction in the rate of polyp recurrence associated with vitamin supplementation is small, and a larger study would be required to ensure that an effect of this size was not a chance finding.