A randomized, prospective study was conducted between October, 1994 and April, 1996 to compare the efficacy and safety of warfarin (group I) with dextran (group II) for the prevention of DVT after hip arthroplasty. 109 cases in 106 patients were included in the study. There were 60 cases in group I and 49 cases in group IL Risk factor for DVT were analysed before surgery and prophylactic modalities were performed. Color doppler was performed to all patients by one muskuloskeletal radiologist on 5th 8th post operative day and on 6 weeks after operation. DVT occurred in ten ( 16.6% ) of the 60 patients in group I, six ( 12.2% ) of the 49 patients in group II. but, this difference was not significant ( p= 0.582 ). Statistical analysis for the risk factors of age, sex, operation time, anesthetic method and operation method were performed and there were no statistical difference. Only statistical significance was a higher rates of DVT in cemented hip arthroplasty (p=0.028). There were seven proximal thigh DVT and nine calf DVT. We confirmed with sequential color doppler image that all cases of DVT except two complicated cases were resolved within 8 weeks after treatment: one case was expired for the reason of PE, and the other case had thigh hematoma during the treatment with warfarin. We could detect only one case of late onset DVT.