Abstract
In this paper, we consider one of the decisions that have to be made about a screening programme: which type of test to use. Our study shows that knowledge of the sensitivities, specificities, and costs of alternative tests is an inadequate basis for the choice of test. The monetary values of the different possible results of the test must also be estimated, or judgements made about the likely magnitude of these values. If judgements have to be made, they should be explicit, because different individuals are likely to judge differently, and their opinions will critically affect the choice of test.