Activated protein C (APC) resistance is a common risk factor for venous thromboembolism and is associated with the replacement of Arg 506 by Gln in the factor V gene (factor V Leiden). We investigated the risk of recurrence of venous thromboembolism in APC resistant patients heterozygous for FV Leiden and compared these patient groups with a group of patients, who had a history of venous thromboembolism, but had neither APC resistance nor the FV Leiden mutation. APC resistance was determined in frozen blood samples from patients with a history of venous thromboembolism, who were not receiving oral anticoagulant (OAC) treatment. The plasma samples were collected between 1984 and 1991. Twenty-one patients in whom APC resistance was found in the stored plasma samples were reinvestigated in 1994 (5 males, 16 females, median [m] age 49 years, range 21-71 years). Twenty-one sex and age matched patients with venous thromboembolism (5 males, 16 females, age m = 50 years, range 25-73 years) investigated during the same time period who had a normal APC resistance test served as a control group. Patients with APC resistance as well as controls were reinvestigated for the presence of FV Leiden by genetic analysis in 1994. Of the 21 APC resistant patients, 5 were homozygous and 16 heterozygous for FV Leiden. Before the study entry homozygous patients had a significantly higher recurrence rate (5/5 patients) compared to the control group in heterozygous patients (9/16) and controls (9/21) the recurrence rate was not significantly different. The total observation time was 21 years in patients with homozygous FV Leiden, 83 years in patients with heterozygous FV Leiden and 108 years in controls, excluding the time when patients were on OAC treatment. During the observation time the recurrence rate was highest in patients with homozygous FV Leiden (9.5% per patient per year), but was similar in patients with heterozygous FV Leiden (4.8% per patient per year) and controls (5% per patient per year). Two of five (40%) homozygous patients, 4/16 (25%) heterozygous and 5/21 (24%) controls had a least one recurrent event during the observation period. The probability for development of thrombosis in the Kaplan-Meyer-Plot analysis was not different between the three groups. Bearing limitations of our study in mind (retrospective design, relatively small patient number) we conclude that the risk of recurrence after a thromboembolic event is not higher in patients with heterozygous FV Leiden than in patients without this mutation. Thus, it appears that the identification of heterozygous FV Leiden mutation is not an indication for long-term OAC treatment. Also, long-term OAC treatment cannot generally be recommended for homozygous patients with a single thromboembolic event. More definitive conclusions will require larger prospective studies.