In order to study whether a low molecular weight heparin (LMWH) of mw 4000 D is effective in the treatment of deep venous thrombosis (DVT), patients with DVT verified by phlebography were randomized to treatment by continuous intravenous infusion of either unfractionated heparin (UFH) or LMWH. The initial dose was 240 U (anti F Xa)/kg/12 h. This study (study I) was stopped because of major bleeding in 2 newly operated patients in the LMWH group after 27 patients had been treated. The heparin activity measured as F Xa inhibition assayed in retrospect, was found to be much higher in the LMWH group (mean 1.6-2.0 anti F Xa U/ml) than in the UFH group (mean 0.5-0.8 anti F Xa U/ml). A second study was therefore initiated in which the DVT patients were randomly given UFH (240 U/kg/12 h) or LMWH only 120 U (anti F Xa)/kg/12 h, as initial doses (study II). In this study 27 patients could be evaluated, the mean heparin activity still being higher in the LMWH group (0.9-1.2 anti F Xa U/ml) than in the UFH group (0.5-0.7 anti F Xa U/ml). A second phlebographic investigation showed progression of thrombus size in 3 (11%) of the UFH patients of studies I and II (n = 29) and improvement in 14 (48%). There was no progression in any LMWH patient, 6 (50%) had improved in study I and 10 (77%) in study II. The mean decrease of thrombus size score (according to Marder) during treatment did not differ between the 3 groups.(ABSTRACT TRUNCATED AT 250 WORDS)