BACKGROUND Identifying the etiologic microorganism is essential to guide antimicrobial therapy in prosthetic joint infection. OBJECTIVE We (1) compared the frequency of positive cultures with synovial fluid inoculated in blood culture flasks (SF) with those of periprosthetic tissues or swabs in traditional cultures from patients with acute and chronic prosthetic joint infections (PJI) and (2) determined the sensitivity, specificity, and predictive values of the three methods. METHODS We retrospectively reviewed 87 patients with PJIs (54 knees, 33 hips) and 63 patients with aseptic loosening (34 knees, 29 hips). Two SF, periprosthetic tissue, and swab samples were taken for culture in all 150 patients except for 14 in whom only one SF fluid sample was obtained. Synovial fluid was inoculated in blood culture flasks and periprosthetic tissue and swab samples in standard media. Positive cultures were identified with standard biochemical procedures. RESULTS SF samples were positive in 78 of 87 infected cases (90%), periprosthetic tissue samples were positive in 71 (82%), and swab samples were positive in 59 (68%). SF, periprosthetic tissue, and swab samples were positive more frequently in acute than in chronic infections (96% versus 82% for SF, 87% versus 74% for periprosthetic tissue, and 87% versus 44% for swabs). The sensitivity, specificity, and positive and negative predictive values of SF were 91, 100, 100, and 93 for acute infections and 79, 100, 100, and 88 for chronic infections, respectively. CONCLUSIONS SF samples cultured in flasks had higher sensitivity, specificity, and positive and negative predictive values for diagnosis of PJI when compared with standard tissue and swab samples. The usefulness of all samples was less in chronic than in acute infections. METHODS Level II, diagnostic study. See Guidelines for Authors for a complete description of levels of evidence.