METHODS: A total of 78 people with medial knee OA (mean ± SD age 62.1 ± 6.9 years, mean ± SD body mass index 29.4 ± 4.0 kg/m(2) , and radiographic disease severity 19% mild, 49% moderate, and 32% severe) who completed a 12-week physiotherapist-supervised exercise trial were randomly allocated to 2 30-minute physiotherapy booster sessions (delivered by 8 physiotherapists in private clinics) or no booster sessions for the subsequent 24 weeks. All participants were asked to continue home exercises 4 times weekly. Primary outcomes were change in pain, using a 100-mm visual analog scale, and self-reported physical function, measured using the Western Ontario McMaster Universities Osteoarthritis Index. Participants and physiotherapists were unblinded to group allocation, although participants were blinded to the study hypothesis.