Rationale The aim of the treatment of osteoarthritis (OA) is to improve symptoms and function; ideally, treatment would also prevent the structural progression of the disease. The targets of structure-modifying OA drugs (STMOADs) can be one of the various lesions of OA, including cartilage destruction, osteophytes, bone sclerosis, cysts, and bone attrition. However, prevention of cartilage destruction (chondroprotection) is currently accepted as the most promising target of STMOADs. The possibility of preventing progression of cartilage damage has been demonstrated in various animal models of OA. Chondroprotection is now being investigated in human OA. Several trials are in progress, and two positive studies have already been reported 1,2 . The goal of these initial trials was mainly to demonstrate that a reduction of the natural progression of OA joint space narrowing is possible in humans. Today, chondroprotection is no longer considered a myth, and the questions now are whether STMOADs provide a clinical benefit for the patient and how to evaluate that benefit. These new and relevant questions are difficult to answer. Can we expect a clinical benefit from STMOADs? Drugs evaluated for a structure-modifying effect may or may not also have symptomatic effects in the short term. STMOADs with a symptomatic effect can be expected to lead to clinical improvement in the OA patient in the long term. Evaluation of the clinical benefit of STMOADs could be possible with the tools presently used for the assessment of symptom-modifying drugs in OA. The clinical benefit that can be expected from a drug without known symptomatic effects but with a structuremodifying effect and how to measure that clinical benefit are as yet unknown. An effective STMOAD is not expected to induce a repair of the structural lesions of OA, but it is expected to stop or decrease their progression. Such a structural effect in the long term might result in symptom improvement or might only prevent a progression of symptoms. It also might have no clinical effect, since symptoms and lesions are poorly correlated in OA patients. The time required to demonstrate the clinical benefit of a STMOAD is unknown. It is possible that there may be a long delay (several years) between the beginning of STMOAD efficacy and the realization of any clinical benefit. The extent of clinical benefit could be related to the degree of effectiveness of the STMOAD. If this were true, the delay in clinical benefit could be very long for drugs able to only partially reduce the structural progression of the disease. Finally, the clinical benefit of a STMOAD might be missed by our usual tools and could require new methods of evaluation.