This prospective study compared PCR and culture techniques in the diagnosis of prosthetic joint infection (PJI). We obtained joint fluid samples (JFS; n=115) from patients who had failed total joint arthroplasty between January 2003 and June 2005; 49 were positive for PJI according to established strict criteria. JFS were analyzed by PCR (n=35; control n=66) or culture (n=46, control n=48). PCR was positive in 71% of PJI cases, resulting in sensitivity, specificity, accuracy, positive predictive value, negative predictive value, and likelihood ratio for positive results as follows: 0.71; 0.97; 0.88; 0.93; 0.87 and 23.6, respectively. Culture was positive in 44% of PJI samples. Corresponding statistics were 0.44; 0.94; 0.69; 0.87; 0.63 and 7.0, respectively. Significantly higher sensitivity, accuracy and negative predictive values were calculated for PCR versus culture, and there was 83% concordance between the results of intraoperative culture and PCR detection of causative bacteria. Therefore, we conclude that PCR analysis of synovial fluid increases the utility of pre-operative aspiration for patients who require revision total joint surgery.