OBJECTIVE: This study seeks to determine the costs and effectiveness of warfarin, ardeparin, enoxaparin, and no-prophylaxis for patients undergoing knee arthroplasty. METHODS: A decision–analytic model was developed using Tree-Age software to compare the three treatment strategies and a no-prophylaxis option. The effectiveness measure was deaths prevented for a simulated cohort of 10,000 patients undergoing knee arthroplasty. Costs were valued in U.S. dollars using a payers perspective and costs and probabilities of events were obtained from the published literature. RESULTS: All three treatment strategies were cost saving and more effective than the no prophylaxis option. Enoxaparin had the lowest expected cost of $3,242 per patient and prevented 198 deaths for the cohort of 10,000 persons. Warfarin was dominated by enoxaparin and ardeparin was the most effective option and had a marginal cost effectiveness ratio of $207,342 per death avoided. Results of the sensitivity analysis will be presented. CONCLUSION: All three treatments are cost saving and more effective than no prophylaxis indicating that prophylaxis is preferred to the do nothing strategy. Warfarin was dominated by enoxaparin and should not be considered a first line anticoagulant to prevent DVT in this patient population. Ardeparin was the most effective option which had a marginal cost-effectiveness ratio above many societal willingness to pay thresholds and may not be considered a desirable use of health care resources. The robustness of these findings will be explored using sensitivity analysis.