OBJECTIVES: Post-discharge thromboprophylaxis is the practice of prescribing antithrombotic therapy for 21 days after discharge, commonly used in surgical patients who are at high risk for venothromboembolism (VTE). While randomized controlled trials have demonstrated a risk reduction for VTE after major general surgery, the incidence rate where it is cost effective has not been established.  Previous cost analyses have not included an effectiveness component, and have not reported a threshold VTE incidence rate to help answer for which procedures it should be implemented. This study sought to determine the VTE incidence threshold for the cost-effectiveness of low molecular weight heparin for 4 weeks after surgery as compared to inpatient prophylaxis only. METHODS: A cost-effectiveness decision tree was created using TreeAge.  Assigned probabilities were derived from published literature.  The decision point compared extended duration thromboprophylaxis with low molecular weight heparin for 21 days after discharge to inpatient-prophylaxis alone, with base case assumptions based on an abdominal oncologic resection without complications in a 45 year-old male.   The end points were pulmonary embolism or deep vein thrombosis with attendant costs and assigned effectiveness evaluated by QALY.  Willingness to pay was set at $50,000/QALY.  Sensitivity analyses were performed to assess uncertainty within the model, with particular interest in the threshold for cost-effectiveness based on VTE incidence. RESULTS: Given base case assumptions with VTE probability of 4%, extended duration thromboprophylaxis had an incremental cost effectiveness ratio of  $8123/QALY, which was considered cost-effective.  The results were robust to sensitivity analysis with the highest uncertainty associated with VTE incidence and medication cost.  The threshold for the relative cost-effectiveness was a VTE incidence exceeding 2.53%.  CONCLUSIONS: Given the base case assumptions, extended prophylaxis is more cost effective than inpatient prophylaxis alone, and the threshold for its use should be cases where the estimated VTE risk exceeds 2.53%.