A simple in vitro test to calculate the dose of heparin required to achieve optimal in vivo anticoagulation during surgery has been assessed in 15 patients who subsequently underwent vascular surgery. Heparin was added to four aliquots of patients blood in vitro to give five solutions with heparin concentrations ranging from 0-0.8 units/ml of plasma. The activated partial thromboblastic times (APTT) of each of these samples was then measured and the natural log (ln) of the APTT calculated. The natural log of the APTT in vitro was then plotted against the in vitro heparin concentration. From this linear correlation the concentration of heparin required to achieve an APTT 2.5 times the normal in vitro (Hc) for the 15 different patients was calculated and ranged from 0.4-0.75 units/ml (median 0.47). Based on an estimate of the plasma volume (PV), the bolus dose of heparin given intravenously to each patient to produce an equivalent anticoagulant response in vivo was calculated (Hc x PB). Heparin boli administered ranged from 1000-2000 units (median 1500). The mean in vivo APTT achieved was 77% of the predicted value (range 62%-123%). Such an estimation of an in vivo response, by means of an in vitro test, should help to more accurately predict the effects of heparin in vivo and individualise anticoagulation dosage.