Aims. Unicompartmental knee arthroplasty (UKA) is the preferred treatment for anterior medial knee osteoarthritis (OA) owing to the rapid postoperative recovery. However, the risk factors for UKA failure remain controversial. Methods. The clinical data of Oxford mobile-bearing UKAs performed between 2011 and 2017 with a minimum follow-up of five years were retrospectively analyzed. Demographic, surgical, and follow-up data were collected. The Cox proportional hazards model was used to identify the risk factors that contribute to UKA failure. Kaplan-Meier survival was used to compare the effect of the prosthesis position on UKA survival. Results. A total of 407 patients who underwent UKA were included in the study. The mean age of patients was 61.8 years, and the mean follow-up period of the patients was 91.7 months. The mean Knee Society Score (KSS) preoperatively and at the last follow-up were 64.2 and 89.7, respectively (p = 0.001). Overall, 28 patients (6.9%) with UKA underwent revision due to prosthesis loosening (16 patients), dislocation (eight patients), and persistent pain (four patients). Cox proportional hazards model analysis identified malposition of the prostheses as a high-risk factor for UKA failure (p = 0.007). Kaplan-Meier analysis revealed that the five-year survival rate of the group with malposition was 85.1%, which was significantly lower than that of the group with normal position (96.2%; p < 0.001). Conclusion. UKA constitutes an effective method for treating anteromedial knee OA, with an excellent five-year survival rate. Aseptic loosening caused by prosthesis malposition was identified as the main cause of UKA failure. Surgeons should pay close attention to prevent the potential occurrence of this problem. Cite this article: Bone Jt Open 2023;4(12):914–922