Objective: It is suggested that patients with acromegaly have an increased risk of colorectal cancer and pre-malignant adenomatous polyps. However, the optimum frequency with which colonoscopic screening should be offered remains unclear. Design: To determine the optimum frequency for repeated colonoscopic surveillance of acromegalic patients. Methods: We retrospectively reviewed the case records of all patients with acromegaly seen in our centre since 1992: 254 patients had at least one surveillance colonoscopy, 156 patients had a second surveillance colonoscopy, 60 patients had a third surveillance colonoscopy and 15 patients had a fourth surveillance colonoscopy. Results: The presence of hyperplastic or adenomatous polyps was assessed in all patients, while one cancer was detected at the second surveillance. At the third surveillance, mean (GS.D.) serum IGF1 levels (ng/ml) in patients with hyperplastic polyps were significantly higher than those with normal colons (P!0.05). The presence of an adenoma rather than a normal colon at the first colonoscopy was associated with a significantly increased risk of adenoma at the second (odds ratio (OR) 4.4, 95% confidence interval (CI) 1.9-10.4) and at the third (OR 8.8, 95% CI 2.9-26.5) screens. Conversely, a normal colon at the first surveillance gave a high chance of normal findings at the second (78%) or third surveillance (78%), and a normal colon at the second colonoscopy was associated with normality at the third colonoscopy (81%). Conclusions: Repeated colonoscopic screening of patients with acromegaly demonstrated a high prevalence of new adenomatous and hyperplastic colonic polyps, dependent on both the occurrence of previous polyps and elevated IGF1 levels.