Bacteriological sampling in orthopedic revision surgery for arthroplasty or internal fixation raises several questions. 1) When? And should sampling be systematic? Sampling should not be systematic in revision surgery, but only in case of suspected infection, in which case empirical antibiotic regimen should be systematically implemented. 2) How? Which tissues, how many and what transport? Only deep samples, preferably taken without ongoing antibiotic therapy, allow reliable interpretation of results. The optimal number of intra-operative samples is 5, or 3 if the laboratory uses seeding in aerobic and anaerobic vials. Samples should be transported to the laboratory within 2 h, at room temperature. 3) What conclusions can be drawn, using what references? There are several classifications, leading to divergent interpretation. The EBJIS (European Bone and Joint Infection Society) classification showed the best sensitivity in a multicenter study. 4) What duration of antibiotic washout before revision, and how to proceed if it cannot be achieved? The antibiotic-free period before sampling should be 14 days, or 21 days in case of prior treatment by cyclins, clindamycin, rifampicin or drugs with a very long half-life such as lipoglycopeptides, except when surgical intervention is required urgently. 5) How to deal with microbiological sampling and antibiotic prophylaxis at the time of revision surgery? Pursuing prophylactic antibiotic therapy during bone and joint implant revision does not greatly impair the value of intra-operative sampling. However, evidence of benefit of continuing antibiotic prophylaxis during revision arthroplasty is lacking. 6) What samples for atypic infection? Atypic micro-organisms (mycobacteria, fungi, etc.) require specific screening, guided by the clinical context and discussed before sampling is carried out.

V; expert opinion.