Abstract Purpose Medial opening wedge high tibial osteotomy (MOWHTO) is an effective procedure for managing isolated medial compartment osteoarthritis (OA) with varus malalignment. This study investigates the effect of radiographic OA severity on the clinical outcomes and survivorship of MOWHTO. Methods A retrospective analysis of a prospectively maintained single‐centre database of 1170 knee osteotomies, between 2002 and 2022, was conducted. Adult patients who underwent MOWHTO for symptomatic unicompartmental medial knee OA, with varus knee malalignment, who had failed conservative management were included. Patients were stratified into four groups according to the radiographic OA severity of the medial compartment based on the Kellgren Lawrence (KL) grading system. Multiple patient‐reported outcome measures (PROMs) were recorded preoperatively and at 2 and 5 years postoperatively. The delta values between preoperative and 2 and 5‐year postoperative PROMs scores were calculated and compared with their reported minimal clinically important difference (MCID) scores. Deformity analysis was undertaken preoperatively and postoperatively. The rate of conversion to arthroplasty, and 5‐ and 10‐year survivorship were recorded, as well as the hazard ratio (HR) of OA KL grade on survivorship. Results A total of 605 cases were included in the study: KL1 group ( n = 71), KL2 group ( n = 203), KL3 group ( n = 210) and KL4 group ( n = 121). There was no intergroup significant difference in the demographics. The mean follow‐up for the whole series was 13.2 ± 3.8 years. There was an intergroup significant difference in the mechanical tibiofemoral angle (mTFA), joint line convergence angle (JLCA), and Mikulicz point (all p < 0.001). A more varus TFA was noted in patients with more advanced OA grades: KL1 −4.8° ± 2.7°, KL2 −5.5° ± 2.9°, KL3 −5.7° ± 3.1° and KL4 −6.8° ± 3.1°. Similarly, there was a lower Mikulicz percentage in higher KL grades: KL1 25.5 ± 12%, KL2 25.0 ± 12.7%, KL3 21.9 ± 13.2% and KL4 17.7 ± 13.9%. Clinically, there were significant improvements across all PROMs and achievement of MCID in KOOS, OKS and VAS pain in all groups. No significant difference in survival outcomes was observed between the four OA KL grade groups ( p = 0.8). Conclusion MOWHTO was associated with durable survivorship and excellent patient‐reported outcomes across the spectrum of radiographic OA severity. However, the extent to which MOWHTO may delay the need for total knee arthroplasty requires confirmation through longer term, prospective studies. Symptomatic patients with early‐stage OA (Kellgren–Lawrence Grade 1) should be counselled on the potential for more limited functional gains following the procedure. Level of Evidence Level IV, retrospective cohort study.