The effect of LMW heparin (Kabi 2165, Fragmin) was compared with placebo for the prevention of DVT in 103 patients with acute ischemic stroke using a prospective, double-blind, randomized trial design. Treatment was started within 72 hours, and LMW heparin was administered subcutaneously once daily according to body weight classes, which corresponded to about 55 to 65 Factor-Xa inhibitory U/kg, for 14 days, or until discharge from the hospital, if earlier. All patients underwent thrombosis surveillance with unilateral venography of the paretic limb. Evaluation of venography could be performed in 42 of 52 patients randomized to LMW heparin and in 50 of 51 patients randomized to placebo. The frequency of DVT was 15 of 42 patients or 36% (95% confidence interval 22 to 52%) in the LMW heparin group and 17 of 50 patients or 34% (21 to 49%) in the placebo group. The frequency of proximal thrombi was 5 of 42 (12%) and 8 of 50 (16%), respectively. There was one fatal pulmonary embolism in the placebo group. The mortality rate (28 days follow-up) was 5 of 52 in the LMW heparin group and 1 of 51 in the placebo group (p = 0.24). None of the deaths was related to treatment. No major hemorrhagic complications were observed. The mean Factor Xa inhibitory activity levels at peak concentration were 0.34 U/ml on day 2 and 0.42 U/ml on day 12 (p = 0.02). We conclude that LMW heparin in the dose range studied did not provide efficient prophylaxis against DVT in patients with acute ischemic stroke.