The semiquantitated immunochemical fecal occult blood test (I-FOBT) used for colorectal cancer (CRC) screening has had its long-term performance characteristics determined by clinical follow-up or sometimes by colonoscopy as the 'gold standard'. We reanalyzed a file of total colonoscopy patients who also prepared three I-FOBTs, processed by the OC-MICRO instrument, using at least 50 ng Hb/ml buffer threshold to determine a positive test. The performance of both tests was evaluated by the National Cancer Registry follow-up to identify new CRCs and by determining the effects of the number of tests prepared and their thresholds for analysis, sex, and age on results. A total of 1630 patients, mean age 62.7 years, SD 11.9, 50.1% men, having undergone both tests were followed up for a mean of 51.5 months, SD 13.4; 25 CRC patients were registered. At 36 months, I-FOBT sensitivity for CRC was 95.8% (95% confidence interval 87.8, 104), as was initial colonoscopy; within 48 months, it was 92% (95% confidence interval 81.4, 103) and 96%, respectively. I-FOBT identified 70 of the 122 (57.2%) colonoscopy-detected advanced adenoma patients. CRC and advanced adenomatous polyps were more common in men (P<0.01), whose risk increased at 51-73 years (odds ratio 4.639, P=0.056), but not among women (odds ratio 1.952). It then increased significantly (P<0.01) for both sexes aged at least 74 years. I-FOBTs identified most CRCs diagnosed within 36 months of follow-up with sensitivity similar to that of initial colonoscopy, with neither test identifying every CRC patient. Sex and age influence results and need consideration when planning population screening.