The argument for laboratory control of heparin treatment rests on 2 issues: 1, that the response to heparin varies from patient to patient and from time to time, and 2, that the clinical results are improved by laboratory control.The theoretical basis for lab control is strong. AT III levels vary and are reduced in thrombotic states. Heparin’s absorption and rate of degradation vary; its survival is shorter after pulmonary embolism.The clinical evidence that lab control prevents bleeding is weak. When heparin is given by intermittent injection, bleeding is not avoided by lab tests. Even with continuous infusion, coagulation tests may identify a group at high risk for bleeding, but lower values are no guarantee against hemorrhage.The case for lab control to assure an anti thrombotic effect is stronger. Clotting test values correlate with recurrent thromboembolism, particularly when heparin is given by continuous infusion. Thromboembolism recurs more ofter after smaller doses or after a poor APTT response to a larger dose. Other clinical settings are analogous: e.g., cardiopulmonary bypass and hemodialysis. New tests may correlate better with clinical events by measuring end products of heparin’s antithrombotic activity rather than its action on coagulation kinetics: e.g., fibrinopeptide A, fibrin monower, thrombin-antithrombin complexes, prothrombin fragments, and products of platelet activation.