Background: Peri-prosthetic joint infection (PJI) is a devastating complication of joint replacement surgery. Determining the optimal duration of intravenous (IV) antibiotics for PJI managed with debridement and implant retention (DAIR) is a research priority. Methods: Patients undergoing DAIR for early and late-acute PJI of the hip or knee were randomised to receive 2 (short-course) or 6 (standard-course) weeks of IV antibiotics, with both groups completing 12 weeks of antibiotics in total. The primary endpoint of this pilot, open-label, randomised trial was a 7-point ordinal desirability of outcome ranking (DOOR) score, which accounted for mortality, clinical cure and treatment adverse events at 12 months. Duration of IV treatment was used as a tiebreaker, with shorter courses ranked higher. Outcome adjudication was performed by expert clinicians blinded to the allocated intervention (Australia and New Zealand Clinical Trials Registry ACTRN12617000127303). Results: 60 patients were recruited; 31 and 29 were allocated to short- and standard-course treatment, respectively. All had an evaluable outcome at 12 months and were analysed by intention-to-treat. Clinical cure was demonstrated in 44 (73%) overall; 22 (71%) in the short-course group and 22 (76%) in the standard-care group (P=0.77). Using the DOOR approach, the probability that short- was better than standard-course treatment was 59.7% (95% confidence interval 45.1-74.3). Conclusions: In selected patients with early and late-acute PJI managed with DAIR, shorter courses of IV antibiotics may be appropriate. Due to small sample size, these data accord with, but do not confirm, results from other international trials of early transition to oral antibiotics.