Cancer of the colon and rectum is a major cause of ill-health. Options for reducing the burden of the disease include primary prevention, screening for early stage asymptomatic disease and improvements in the treatment of symptomatic disease. If the policy objective is to make a major impact on mortality from the disease then screening appears to be the only technically feasible option. One indication of asymptomatic colorectal cancer is small quantities of blood mixed with faeces. Screening tests capable of detecting bleeding are currently being evaluated in clinical trials. Interim measures of the costs and disease yield of a screening programme using a faecal occult blood test imply that screening may offer good value for money but only if the intended mortality reduction from the disease is realised. There are various ways of 'fine-tuning' the screening programme to improve the balance of costs and benefits; information for making choices regarding important parameters such as the age range of the population to be offered screening are presented. Alternative screening tests are also evaluated and the results presented in terms of the cost-yield trade-off. The policy implications of the evaluation must be qualified at this stage since no proof of mortality reduction will be available until the conclusion of the ongoing trial. Nevertheless, under various assumptions about the impact of screening, the option appears to be an efficient way of reducing the health 'costs' of colorectal cancer.