INTRODUCTION There is anecdotal evidence of poorer outcomes from prosthetic joint replacement performed in independent treatment centres but very few comparative studies. PATIENTS AND METHODS We compared the early re-operation, dislocation, re-admission, major wound infection and audit rates of 880 total hip (THA) and 874 total knee (TKA) arthroplasties carried out at a regional orthopaedic hospital (Avon Orthopaedic Centre; AOC) with 368 THAs and 365 TKAs from an NHS (WGH) and 67 THAs and 86 TKAs from a private hospital (CNH) independent treatment centre. RESULTS Early re-operation rates were 9% at CNH, 1.4% at WGH and 0.6% at AOC after THA and 8% at CNH, 1.9% at WGH and 1% at AOC after TKA. After THA, dislocation rates after THA were 6% at CNH and 1.8% at both WGH and AOC. Readmission rates were 13% at CNH, 0.6% at WGH and 1.2% at AOC. Major wound problems were 20% at CNH, 3.8% at WGH and 0.4% at AOC after THA and 19% at CNH, 1.9% at WGH and AOC (1.1%) after TKA. After TKA, re-admission rates from CNH were 13%, 1% at WGH and 1.1% at AOC. AOC and WGH audited their outcomes. None were available from CNH. CONCLUSIONS Results and audit from independent treatment centres are variable and patients should be warned of this before undergoing treatment at them.