68 adults with symptomatic medial compartment knee osteoarthritis were randomized to receive either personalised gait retraining or sham gait retraining (34 patients in each arm). All randomized participants were included in the intention-to-treat analysis. The primary outcomes of interest were change in medial knee pain measured using a numeric rating scale and change in knee adduction moment peak during walking. Secondary outcomes included MRI-based cartilage changes assessed using T1ρ and T2 relaxation times in the femoral cartilage. Outcomes were assessed at baseline and at one year. Overall, the results of the study revealed that personalised gait retraining produced significantly greater reductions in medial knee pain and knee adduction moment compared with sham gait retraining and was associated with slowed medial femoral cartilage degeneration on quantitative MRI. These findings suggest that personalised gait retraining is an effective non-surgical intervention for selected patients with medial compartment knee osteoarthritis.