Non-pharmacological interventions are frequently and widely used in the management of patients with osteoarthritis (OA). However, there is little evidence that most of these interventions are efficacious (i.e. that they work under ideal circumstances) owing to the paucity of research studies on these interventions and the fundamental methodological flaws in published studies. Moreover, the clinical effectiveness of non-pharmacological interventions (i.e. their efficacy under the conditions prevailing in clinical practice) is unknown, and cost-analysis of these interventions has not been carried out. If evidence-based management guidelines were to be constructed solely from firm research studies few could be recommended. Established and developing research methodologies and techniques should be employed to construct and conduct research programmes that appropriately evaluate the clinical and cost effectiveness of these complex non-pharmacological healthcare interventions. Establishing clinical effectiveness will ensure that we have practice-based evidence and are administering relevant, effective and optimal healthcare for patients with OA, maximizing the efficient use of healthcare resources.