The thrombotic risk of women with a heterozygous natural clotting inhibitor deficiency taking oral contraceptives (OC) has not been evaluated. Therefore, a retrospective collaborative controlled cohort-study was carried out in 8 coagulation laboratories and thrombosis units in Austria, Germany and Switzerland. The incidence of thromboembolism in 48 females heterozygous for hereditary type I deficiency of antithrombin III (n = 15), protein C (n = 16) or protein S (n = 17), who had taken OC at least once in their life were compared with that of 48 deficient women, who had never taken OC (controls). Diagnosis of the deficiency state was made in the participating centers. Data on the onset and duration of OC intake and the date and site of thrombotic events were obtained from a questionnaire filled in by the patient or a physician during a visit at a participating center. The observation period in the OC patients was started with onset of OC intake and was terminated when a thromboembolic event had occurred or when OC medication were discontinued. In the patients without OC, the observation period began at an age matched to that of the OC patient and ended when a thromboembolic event had occurred or was continued as long as the corresponding OC patient was on treatment. In AT III-deficient females the probability for thrombosis was significantly higher for patients taking OC compared to the non-OC-patients (Wilcoxon test p = 0.004, Log Rank test p = 0.005).(ABSTRACT TRUNCATED AT 250 WORDS)