This review has focused on the prevalence and risk factors associated with knee and hip osteoarthritis. Risk factors for knee osteoarthritis are obesity and major injury, and knee osteoarthritis probably fits into the generalized osteoarthritis diathesis. Repetitive use, such as in jobs requiring heavy labor and knee bending, probably increases the risk of knee osteoarthritis. Hip osteoarthritis is probably frequently secondary to developmental defects. As Rothman (182) has pointed out in discussing causation, this does not necessarily mean that the same factors do not also contribute to causing hip osteoarthritis. Yet, it appears that, in many cases, developmental defects are severe enough to be sufficient causes of hip osteoarthritis. To delineate other causes, it may be necessary to examine risk factors separately in those with and in those without developmental disease. Although large epidemiologic studies are best able to identify the relative contributions of specific risk factors while controlling for other risk factors, new studies need to focus on important unresolved questions. First, longitudinal studies with comprehensive follow-up using repeated radiographic assessments are needed to identify factors that cause development of disease or the onset of symptoms. Second, cohorts with early and possibly asymptomatic disease need to be followed to determine the causes of progression or regression of disease and the natural history of disease. Such cohorts may include those at high risk of injury such as sports enthusiasts or manual laborers.