We examined 136 consecutive patients with histories of guaiac positive stool examinations, bright red blood per rectum, or hematochezia to determine the value of the double-contrast barium-enema (DC-BE) examination and colonoscopy/proctoscopy in establishing bleeding sites. If examination findings were analyzed in conjunction with findings of visual examination of the anal area, the difference in the respective sensitivities of the two examinations was not statistically significant. If findings at the visual inspection were excluded, the DC-BE examination missed 45 of 155 proved bleeding sites (sensitivity, 71%) and 13 of 35 nonbleeding lesions (sensitivity for all lesions 70%), while colonoscopy missed 13 of 155 bleeding sites (sensitivity, 92%) and seven of 35 nonbleeding lesions (sensitivity for all lesions, 90%). The sensitivity of both methods was similar if all rectal and anal lesions were excluded. Only eight additional lesions (all polyps) were found in 78 patients who had bleeding internal and external hemorrhoids, anal fissures, and anal tears, and seven of these were found in patients who were older than 50 years of age. We conclude that colonoscopy is superior to the DC-BE examination in the detection of bleeding sites but similar in results to the DC-BE examination if lesions in the anal canal and rectum are excluded. The routine use of the DC-BE examination in patients with superficial lesions in the anal canal should be discouraged unless the patient has persistent bleeding or is 50 years old or older.