The efficiency of double contrast barium enema and flexible rectosigmoidoscopy (to 60 cm) in the assessment of patients with a positive Hemoccult II® test in a randomized screening study for colorectal neoplasms was evaluated. A positive test was present in 625 patients, of whom 530 had a complete enema and rectosigmoidoscopic assessment. A carcinoma was diagnosed in 26 and an adenoma ⩾ 1 cm in diameter in 71. As a control, 323 patients with a negative assessment repeated the Hemoccult II test and of these 67 had a positive second test, of whom 55 underwent colonoscopy. One carcinoma (Dukes' A) and two adenomas ⩾ 1 cm in diameter were diagnosed. The efficiency of the assessment was also checked by rescreening the whole group 1–2 years after the first study and by continuing follow-up. It was found that two more carcinomas and one adenoma ⩾ 1 cm in diameter had been overlooked at the primary assessment. The sensitivity for neoplasms ⩾ 1 cm in diameter at the primary assessment was 72 per cent for double contrast barium enema and 86 per cent for rectosigmoidoscopy. The sensitivity for the combined methods was 94 per cent and the specificity was 99 per cent. The combination of double contrast barium enema and rectosigmoidoscopy in the primary assessment of patients with a positive Hemoccult II test gives an acceptable result and immediate retesting of those with a negative assessment is not necessary.