Introduction: When treating periprosthetic joint infections with a two-stage procedure, antibiotic-impregnated spacers are used in the interval between removal of prosthesis and reimplantation. According to our experience, cultures of sonicated spacers are most often negative. The objective of our study was to investigate whether PCR analysis would improve the detection of bacteria in the spacer sonication fluid. Methods: A prospective monocentric study was performed from September 2014 to January 2016. Inclusion criteria were two-stage procedure for prosthetic infection and agreement of the patient to participate in the study. Beside tissues samples and sonication, broad range bacterial PCRs, specific S. aureus PCRs and Unyvero-multiplex PCRs were performed on the sonicated spacer fluid. Results: 30 patients were identified (15 hip, 14 knee and 1 ankle replacements). At reimplantation, cultures of tissue samples and spacer sonication fluid were all negative. Broad range PCRs were all negative. Specific S. aureus PCRs were positive in 5 cases. We had two persistent infections and four cases of infection recurrence were observed, with bacteria different than for the initial infection in three cases. Conclusion: The three different types of PCRs did not detect any bacteria in spacer sonication fluid that was culture-negative. In our study, PCR did not improve the bacterial detection and did not help to predict whether the patient will present a persistent or recurrent infection. Prosthetic 2-stage exchange with short interval and antibiotic-impregnated spacer is an efficient treatment to eradicate infection as both culture- and molecular-based methods were unable to detect bacteria in spacer sonication fluid after reimplantation.