A prospective study was carried out to see whether a small fixed dose of warfarin (1 mg daily) given before operation (mean 20 days) would prevent deep vein thrombosis in patients having major gynaecological surgery. One hundred and four patients were randomised into three groups: fixed minidose warfarin; full dose oral anticoagulation; and no treatment (controls). There was a significantly lower incidence of deep vein thrombosis in the minidose warfarin and full dose anticoagulant treatment groups (9% (3/32) and 3% (1/35) respectively) than in the controls (30%; 11/37) but no significant difference between the two anticoagulant treatment groups. Prothrombin time and the activated partial thromboplastin time were normal on the day of surgery in the warfarin treatment group, whereas times were prolonged in the group given full dose anticoagulation. Mean haemoglobin concentrations fell in all three groups after operation but the fall was significantly less in the minidose warfarin treatment group than after full dose anticoagulation. The benefit from full dose oral anticoagulant prophylaxis, based on a preoperative international normalised ratio of 1.5-2.5 with rabbit brain Manchester reagent, was similar to the protection achieved in an oral anticoagulant treatment group controlled with human brain Manchester comparative reagent at a similar level of anticoagulation. The lack of disturbance of normal haemostasis at the time of operation together with a significant reduction in deep vein thrombosis may encourage surgeons to introduce minidose prophylaxis with warfarin.