BACKGROUND: The efficacy of polypectomy in preventing colorectal cancer (CRC) has never been demonstrated in a controlled, prospective study. This must be done by randomization within a population with a high prevalence of colorectal polyps, and the feasibility and safety of endoscopic screening examination is a prerequisite for this type of study. METHODS: The present study is a randomized, controlled study of the feasibility and safety of flexible sigmoidoscopic screening of a normal population sample of 799 men and women aged 50-59 years, findings at 2 and 6 years' colonoscopic follow-up, and the appearance of clinical colorectal cancer (CRC) after 10 years. RESULTS: The attendance rate was high, and there were no complications. After 10 years 1 of 400 in the screening group had developed CRC (in the group of 76 (19%) not attending for screening examination). Four of 399 controls developed CRC. CONCLUSIONS: Poor yield of polyps at follow-up, slow growth of in situ polyps, and no clinical CRC among screenees after 10 years provides support to infrequent or no colonoscopic follow-up after initial polypectomy in individuals with otherwise average risk of CRC.