BACKGROUND: The sensitivity of unhydrated Haemoccult II has been examined in the context of a randomized controlled trial of faecal occult blood screening for colorectal cancer in Nottingham, UK. METHOD: Both traditional and proportional incidence methods were used to calculate sensitivity separately for both sexes, for two age groups at entry to the trial, for first screen and repeat screens and for three subsites within the large bowel. RESULTS: The traditional method of estimation yielded a sensitivity of 59% whereas the corresponding figure obtained using the proportional incidence method was 54%. The difference between the estimates using the two methods was greatest in subjects aged > or = 65 at entry to the trial and in cancers of the distal colon. CONCLUSIONS: The results suggest that there may be a higher proportion of slower growing tumours in subjects aged > or = 65 and that cancers occurring in the distal colon may have a longer mean sojourn time than cancers proximal to the sigmoid colon.