ObjectivesTo describe the prevalence rates of revision surgery for the treatment of prosthetic joint infection (PJI) for patients undergoing knee replacement, their time trends, the cumulative incidence function of revision for PJI and estimate the burden of PJI at health service level.DesignWe analysed revision knee replacements performed due to a diagnosis of PJI and the linked index procedures recorded in the National Joint Registry from 2003 to 2014 for England and Wales. The cohort analysed consisted of 679 010 index primary knee replacements, 33 920 index revision knee replacements and 8247 revision total knee replacements performed due to a diagnosis of PJI. The prevalence rates, their time trends investigated by time from index surgery to revision for PJI, cumulative incidence functions and the burden of PJI (total procedures) were calculated. Overall linear trends were investigated with log-linear regression.ResultsThe incidence of revision total knee replacement due to PJI at 2 years was 3.2/1000 following primary and 14.4/1000 following revision knee replacement, respectively. The prevalence of revision due to PJI in the 3 months following primary knee replacement has risen by 2.5-fold (95% CI 1.2 to 5.3) from 2005 to 2013 and 7.5-fold (95% CI 1.0 to 56.1) following revision knee replacement. Over 1000 procedures per year are performed as a consequence of knee PJI, an increase of 2.8 from 2005 to 2013. Overall, 75% of revisions were two-stage with an increase in use of single-stage from 7.9% in 2005 to 18.8% in 2014.ConclusionsAlthough the risk of revision due to PJI following knee replacement is low, it is rising, and coupled with the established and further predicted increased incidence of both primary and revision knee replacements, this represents an increasing and substantial treatment burden for orthopaedic service delivery in England and Wales. This has implications for future service design and the funding of individual and specialist centres.