Objective To investigate the effect of rivaroxaban on the risk of bleeding after total knee arthroplasty (TKA). Methods A total of 119 cases undergoing primary TKA because of knee osteoarthritis between June 2009 and May 2011, were randomly divided into the rivaroxaban group (59 cases) and the control group (60 cases). There was no significant difference in gender, age, height, weight, side, disease duration, and grade of osteoarthritis between 2 groups (P gt; 0.05). Thepreoperative preparation and operative procedure of 2 groups were concordant. At 1-14 days after TKA, rivaroxaban 10 mg/d were taken orally in the rivaroxaban group, and placebo were given in the control group. The blood routine examination was performed before operation and at 2 days postoperatively; the total blood loss and hemoglobin (HGB) decrease were calculated according to the formula; the blood loss, postoperative wound drainage, and wound exudate after extubation were recorded to calculate the dominant amount of blood loss; and the bleeding events were recorded within 35 days postoperatively. Results The total blood loss and HGB decrease were (1 198.34 ± 222.06) mL and (33.29 ± 4.99) g/L in the rivaroxaban group and were (1 124.43 ± 261.01) mL and (31.57 ± 6.17) g/L in the control group, showing no significant difference (P gt; 0.05); the postoperative dominant blood loss in the rivaroxaban group [(456.22 ± 133.12) mL] was significantly higher than that in the control group [(354.53 ± 96.71) mL] (t=4.773, P=0.000). The bleeding events occurred in 3 cases (5.1%) of the rivaroxaban group and in 1 case (1.7%) of the control group, showing no significant difference (χ2=1.070, P=0.301). Conclusion Rivaroxaban has some effects on the risk of bleeding after TKA. In general, rivaroxaban is safe.