BACKGROUND, AIMS, AND PATIENTS: In a prospective follow up and intervention study of colorectal polyps, leaving all polyps less than 10 mm in situ for three years, analysis of redetection rate, growth, and new polyp formation was carried out in 116 patients undergoing annual colonoscopy. The findings in relation to growth and new polyp formation were applied to 58 subjects who received placebo. RESULTS: Redetection rate varied from 75-90% for each year, and was highest in the rectum and sigmoid colon. There was no net change in size of all polyps in the placebo group, however, polyps less than 5 mm showed a tendency to net growth, and polyps 5-9 mm a tendency to net regression in size, both for adenomas and hyperplastic polyps. This pattern was verified by computerised image analysis. Patients between 50 and 60 years showed evidence of adenoma size increase compared with the older patients, and the same was true for those with multiple adenomas (four to five) compared with those with a single adenoma. The new adenomas were significantly smaller and 71% were located in the right side of the colon. Patients with multiple adenomas had more new polyps at all the follow up examinations than patients with a single adenoma. One patient developed an invasive colorectal carcinoma, which may be evolved from a previously overlooked polyp. Two polyps, showing intramucosal carcinoma after follow up for three years, were completely removed, as judged by endoscopy and histological examination. CONCLUSIONS: The results show that follow up of unresected colorectal polyps up to 9 mm is safe. The consistency of growth retardation of medium sized polyps suggests extended intervals between the endoscopic follow up examinations, but the increased number of new polyps in the proximal colon indicates total colonoscopy as the examination of choice. The growth retardation of the medium sized polyps may partly explain the discrepancy between the prevalence of polyps and the incidence of colorectal cancer.