OBJECTIVE: To determine whether longer duration of anesthesia predisposes patients undergoing orthopedic surgery to venous thromboembolism (VTE). PATIENTS AND METHODS: We conducted a secondary analysis of a retrospective case-control study that examined risk factors for postoperative VTE in postmenopausal women. We matched women aged 50 years and older with radiographically confirmed postoperative VTE (cases) by age, surgeon, year of surgery, and surgical joint (knee vs hip) with women without postoperative VTE (controls). Duration of anesthesia, operative variables, demographic data, comorbid illnesses, and laboratory data were determined by medical record review. RESULTS: Eighty-eight cases were matched with 181 controls. Duration of anesthesia of 3.5 hours or longer (corresponding to the upper tertile of patients) was strongly associated with postoperative VTE compared with a shorter duration of anesthesia (odds ratio, 3.58; 95% confidence Interval, 2.11-6.16; P < .001). This relationship was maintained after controlling for multiple covariates with propensity score methods, Including type of arthroplasty, route of anesthesia, type of antithrombotic prophylaxis, and surgical approach. In multivariate analysis, the Important predictors of VTE included anesthesia duration of 3.5 hours or longer, type of antithrombotic prophylaxis, revision (vs primary) arthroplasty, and allogeneic blood transfusion. CONCLUSION: We found a marked association between the duration of anesthesia and postoperative VTE in patients undergoing Joint arthroplasty. Although it is possible that unmeasured intraoperative variables account for this relationship, we suggest that duration of anesthesia may be an important risk factor for postoperative VTE after orthopedic surgery. [ABSTRACT FROM AUTHOR]