OBJECTIVE To investigate the current problems and corresponding solutions regarding the use of antibiotic-impregnated cement spacer for the treatment of periprosthetic joint infection (PJI). METHODS A retrospective analysis was made on the clinical data of 27 patients with PJI who underwent two-stage revision with antibiotic-impregnated cement spacer between January 2001 and January 2013. There were 12 males and 15 females, with an average age of 62.7 years (range, 25-81 years). All arthroplasties were unilateral, including 19 hip PJI and 8 knee PJI. The mean duration from primary arthroplasty to PJI was 25 months (range, 3-252 months). After infection was controlled with the antibiotic-impregnated cement spacer combined with systematic antibiotics treatment, two-stage revision was performed. The effectiveness was evaluated. Results One patient died of myocardial infarction at 2 days after surgery. Infection was controlled, and two-stage revision was successfully performed in 19 patients; deep venous thrombosis occurred in 1 of 3 patients who experienced hip spacer fractures, which was cured after conservative management. The spacers were removed and bacteria-sensitive antibiotics was used because of recurrent infections after the first-stage surgery in 7 patients; 3 patients gave up treatment because infection was not controlled, 4 patients received revision after infection was controlled. Twenty-three patients were followed up 1-5 years (mean, 2.3 years). The average Harris hip score and KSS score at 1 years after revision were significantly improved when compared with preoperative ones (P<0.05). In the 8 patients with gram-negative or fungus infection, 7 were found to have recurrent infection after the first-stage surgery; in the 12 patients with gram-positive infection, no recurrent infection was found. Failed treatment was observed in 1 patient with gram-positive and gram-negative infections and 2 with fungus infection, respectively. CONCLUSIONS Antibiotic- impregnated cement spacer has a satisfactory effectiveness for PJI. However, complication of spacer fracture should be noted, especially hip spacers. If the pathogen is gram-negative bacteria or fungus, the implanted spacer may increase the possibility of recurrent infection.