Colorectal cancer screening in ulcerative colitis patients is a commonly accepted practice whose parameters are based on convention and not necessarily on the epidemiology of disease. This review summarizes previously published data from a single university-based surveillance program to estimate important screening parameters. The hazard rate (annual risk) of developing cancer was found to rise exponentially with disease duration and to be approximately 2% at 20 yr of disease and nearly 8% at 30 yr of disease. Older age at symptom onset was a significant risk factor for neoplasia. The lead time between the development of low-grade dysplasia and high-grade dysplasia or cancer was approximately 3 yr. Efficient scheduling of tests (maximum benefit for a given cost) implied that the screening test interval should be inversely proportional to the square root of the hazard rate. As the hazard rate increased with duration of disease, the screening interval should shorten accordingly. An historical cohort study demonstrated that the parameters used in our program were not associated with the anticipated result of reduced colorectal cancer mortality. A surveillance program should be tested that uses a more sensitive criterion for a positive test.