purpose: Although flexible sigmoidoscopy is recommended in patients over age 40 to complement fecal occult blood screening for colorectal neoplasia, the yield of this procedure in asymptomatic, average-risk subjects has varied between studies. In addition, the efficacy of flexible sigmoidoscopy in detecting early curable carcinoma has been the subject of debate. We therefore undertook this study to assess the efficacy, safety, and cost of the procedure for early detection of colorectal neoplasia in asymptomatic subjects. patients and methods: Flexible sigmoidoscopy with a 60-cm scope was performed in 412 asymptomatic veterans (mean age, 63.2 years). Subjects with positive fecal occult blood and those at increased risk for colorectal neoplasia were excluded. Costs were estimated on the basis of Medicare payments. results: A mean length of 56.1 cm of sigmoid and descending solon was examined. A total of 132 polyps were detected in 93 subjects (22.6%). Thirty-five percent of the polyps were located at a distance of greater than 30 cm from the anal verge. Of 122 polyps removed, 26% were 1 cm in size or larger. A total of 77 polyps (63%) were adenomas, and six polyps (5%) were carcinomas. Of the six carcinomas detected (two carcinoma in situ, one Astler Coller stage A, two stage B1, and one stage C1), five were in a localized stage and thus potentially curable. The cost of detecting each potentially curable carcinoma was $47.174. No complications from flexible sigmoidoscopy were noted. conclusion: In conclusion, 60-cm flexible sigmoidoscopy is a safe, high-yield procedure for the initial screening of colorectal neoplasia in asymptomatic, average-risk subjects over the age of 50. Because the carcinomas detected were still in an early stage, such screening may improve survival.