Background While it is accepted accurate identification of infecting organisms is crucial in guiding treatment of periprosthetic joint infection (PJI), there remains no consensus regarding the best method for obtaining cultures. Questions/purposes We compared the yield of intraoperative tissue samples versus swab cultures in diagnosing PJI. Methods Tissue and swab cultures (three each) were collected prospectively during a consecutive series of 156 aseptic and septic revision arthroplasties from October 2011 to April 2012. The tissues and swabs were taken from standardized regions of the joint. After excluding 39 reimplantation procedures, we included 117 cases (74 hip, 43 knee; 30 septic, 87 aseptic) for analysis. We used a modified version of the Musculoskeletal Infection Society criteria for defining PJI, requiring three of five rather than four of six criteria. Tissue and swab cultures from septic and aseptic cases were used to calculate their sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) for identifying PJI. Results Tissue cultures were positive in a higher percentage of septic cases than swab cultures: 28 of 30 (93%) versus 21 of 30 (70%). Tissue cultures were positive in two of 87 aseptic cases (2%), while swab cultures were positive in 10 of 87 (12%). The sensitivity, specificity, PPV, and NPV were 93%, 98%, 93%, and 98%, respectively, for tissue cultures and 70%, 89%, 68%, and 90%, respectively, for swab cultures. Conclusions Tissue cultures demonstrated higher sensitivity, specificity, PPV, and NPV for diagnosing PJI than swab cultures. Swab cultures had more false-negative and false-positive results than tissue cultures. Because swab cultures pose a higher risk of not identifying or incorrectly identifying infecting organisms in PJI, we believe their use in obtaining intraoperative culture specimens should be discouraged.