In order to evaluate the effectiveness of colorectal cancer screening by the faecal occult blood test (FOBT) in Germany a population-based case-control study was conducted in Saarland, a southwestern state of Germany. As cases, we identified 522 persons (244 males, 278 females) who died of colorectal cancer between 1983 and 1986 between the ages of 55 and 75 years. For 163 male cases complete screening histories were retrieved together with up to five age-matched controls who had not died of colorectal cancer identified from the files of the case's referring general practitioner (GP). For 209 female cases screening histories were retrieved from their GPs and gynaecologists as well as for age-matched controls. Individual screening histories were established with emphasis on identifying whether FOBTs were carried out asymptomatically or symptomatically. In the time period 6-36 months prior to diagnosis 13% of the male cases and 14% of the male controls had at least one asymptomatic FOBT with a corresponding matched odds ratio of 0.92 (95% CI: 0.61, 1.75). For the same prediagnostic period 16% of the female cases and 29% of the female controls had at least one asymptomatic FOBT leading to an odds ratio of 0.43 (95% CI: 0.27, 0.68). Thus, for males where participation rates are generally low, no protective effect could be seen but for females where participation rates are higher a clear protective effect is seen. Possibilities for bias need to be taken into consideration when interpreting these results. Organizational measures ensuring a high penetrance of a mass screening programme are seen as a way to elevate efficacy.