The reproducibility of polyp localization and size measurements was tested to evaluate the reliability of observations during follow-up examinations of sigmoidoscopically diagnosed polyps. Our findings suggest that 25% of polyps less than 5 mm in diameter but only 5% of those larger than 5 mm are likely to be missed by a single endoscopic examination. Localization measurements made during insertion and withdrawal of the endoscope on different days of examination diverged proximal to 30 cm from the anal ring during repeat colonoscopy. Otherwise, reproducibility was good, demonstrating that sigmoidoscopy may be a fairly accurate procedure for evaluation of site and size of polyps during follow-up studies.