From individual randomized studies, it is not clear whether a closed suction drainage should be used after total knee arthroplasty. Our meta-analysis compares the clinical outcomes of closed suction drainage with nondrainage after total knee arthroplasty in randomized controlled trials reported between January 1966 and May 2010. Fifteen eligible trials involving 1361 knee incisions (686 knees with closed suction drainage and 675 knees without drainage) satisfied the inclusion criteria for our meta-analysis. The result of the meta-analysis indicates that closed suction drainage reduces the incidence of soft tissue ecchymosis and requirement for dressing reinforcement, but increases the rate of homologous blood transfusion. No significant difference between drainage and nondrainage was observed in the incidence of infection, deep venous thrombosis, or postoperative range of motion.