Implementation studies are recommended to assess the feasibility and effectiveness in real-life of programmes which have been tested in randomized controlled trials (RCTs). We report on an implementation study of two evidence-based exercise and health education programmes for older adults with osteoarthritis (OA) of the knee or hip. Three types of primary health-care providers (n = 18) delivered the OA Knee programme (n = 20) and the OA Hip programme (n = 20), supported by programme manuals and implementation guidelines, in four regions. The outcome measures were pain and mobility. The Knee programme had OA knowledge and self-efficacy as additional outcome measures. Differences in outcome measures and background variables of participants were assessed between the RCTs and the implementation study. Positive effects (P less than 0.05) were found for OA knowledge, pain and self-efficacy in the Knee programme (n = 157), and for pain in the Hip programme (n = 132). No effect was found for mobility. Effect sizes of the RCTs and the present study were comparable. Background variables did not explain the variance in the outcome measures. The outcomes of the previous RCTs and the implementation study were comparable, and indicated the ecological validity of the two programmes. The implications for nationwide dissemination and implementation in The Netherlands are discussed. (Contains 6 tables.)