The non-vitamin K antagonist oral anticoagulants (NOACs), rivaroxaban, apixaban, and dabigatran, have been shown in phase 3 trials to be effective for thromboprophylaxis in patients undergoing elective hip or knee arthroplasty. Results from prior studies suggested that the safety of anticoagulants in such patients was improved if the first postoperative dose was delayed for at least 6 h after surgery. The timing of the first postoperative dose of the NOACs tested in phase 2 studies differed among the three NOACs: dabigatran was started 1 to 4 h postoperatively, whereas rivaroxaban and apixaban were started at least 6 and 12 h, postoperatively, respectively. Our review of the timing of initiation of thromboprophylaxis in randomized trials provides three related lessons. First, clinical trials performed before the NOACs were evaluated demonstrated that delaying the first dose of prophylactic anticoagulation until after major surgery is effective and safe. Second, the optimal timing of the first dose of prophylactic anticoagulation after surgery depends on the dose that is selected. Third, the results of the phase 3 trials with NOACs for thromboprophylaxis support the concept that acceptable efficacy and safety can be achieved when the appropriate first postoperative dose of anticoagulant is delayed for at least 6 h after surgery.