PURPOSE: To compare the efficacy of personalized osteotomies with that of standard osteotomies in treating medial unicompartmental knee osteoarthritis. METHODS: The clinical data of 96 patients who were diagnosed with unicompartmental knee osteoarthritis in our group between 2019 and 2023 were retrospectively analysed on the basis of preoperative and postoperative radiological measurements. The knee injury and osteoarthritis outcome score (KOOS), forgotten joint score (FJS), and Lysholm knee score scale (Lysholm) were used to assess the clinical outcome, and complications were observed and recorded. RESULTS: According to the relevant criteria, 84 of 96 patients were included in this study. All patients were followed for a mean of 31 (range 22-55) months. Fifty-one patients underwent personalized osteotomy procedures, and thirty-three underwent standard osteotomy procedures. The postoperative KOOS Pain (P < 0.0001), KOOS Symptoms (P < 0.0001), KOOS ADL (P < 0.0001), KOOS Sport (P = 0.0023), KOOS QoL (P < 0.0001), Lysholm (P < 0.0001) and FJS (P < 0.0001) scores were higher than those in the standard osteotomy group. Nevertheless, postoperative extension (P = 0.2636) and postoperative flexion (P = 0.3554) were not significantly different. CONCLUSION: This was a single-centre, retrospective, short follow-up study with several limitations. However, on the basis of the results of the present study, we believe that the function of the knee after medial unicompartmental knee arthroplasty (mUKA) is affected by the direction of tibial osteotomy. We believe that better clinical results may be obtained when the tibial implant is placed near the preoperative tibial deformity. LEVEL OF EVIDENCE: Level IV; retrospective case series.