Replication of similar findings in diverse study populations using different study designs provides important evidence for causal inferences regarding risk factors reported in epidemiologic studies. Risk factors for knee osteoarthritis have been substantiated in a number of cross-sectional and longitudinal cohort studies. Longitudinal cohort studies have also added to our understanding of risk factors associated with hip osteoarthritis and hand osteoarthritis. In contrast, studies of hormonal and reproductive factors and the risk of developing rheumatoid arthritis continue to report conflicting results, and the importance of these risk factors remains unresolved. Recent studies suggest that the protective effect is limited only to the development of definite rheumatoid arthritis with a more severe clinical course. Cross-sectional studies of diverse populations show consistent age- and gender-specific differences in the rates of hip fracture among persons of Japanese ancestry and rates among several different groups of whites in the United States. The results of a population-based longitudinal cohort study suggest that women who live in communities where the drinking water has a high content of fluoride have a greater risk of osteoporosis and subsequent fractures than those living in communities with low fluoride in the drinking water. Finally, several recent reports provide additional evidence of the increasing incidence of acute rheumatic fever in the United States and other parts of the world.