Concerned by a perceived high revision rate, we retrospectively reviewed the survivorship of a series of 43 cemented, medial, mobile-bearing Preservation unicompartmental knee replacements implanted during a 2-year period at a single institution. The initial post-operative AP and lateral radiographs were independently assessed to test the hypothesis that suboptimal implantation of the prosthesis was responsible for early failure. An X-ray scoring system based on the criteria for assessing the Oxford mobile-bearing unicompartmental knee replacement was devised. The components of this score included assessment of prosthesis alignment, sizing and cementation. Nine (21%) LCS Preservation mobile-bearings prostheses had required revision at a mean of 22 months post-implantation. The commonest causes for failure were pain (44%) and tibial component loosening (33%). Analysis of post-operative radiographs showed no difference (n.s.) between the compound error scores for the revised and the surviving prostheses. No particular surgical error was identifiable leading to subsequent need for revision. The high failure rates shown in this study have led us to cease using this implant. The clinical relevance of this study is that the captive running track of the LCS mobile-bearing prosthesis may over constrain the meniscal component leading to early failure.