The clinical diagnosis of deep venous thrombosis is non-specific, i.e. suffers from a large number of false-positive diagnoses. Therefore, the use of objective tests is emphasized. We have investigated retrospectively the increase, if any, in the use of objective tests over a three year period in the Leiden area (1986-1989). We found that the percentage of patients on whom objective tests were used was more than doubled, from 21% to 55%; this increase was accompanied by a 29% decrease in the incidence of venous thromboembolic disease. These opposing trends are to be expected if more objective tests are used.