The authors used data from the United States first national Health and Nutrition Examination Survey of 1971-1975 (HANES I) to explore the cross-sectional associations between radiographic osteoarthritis of the knee and a variety of putative risk factors. A total of 5,193 black and white study participants aged 35-74 years, 315 of whom had x-ray-diagnosed osteoarthritis of the knee, were available for analysis. After controlling for confounders, the authors found significant associations of knee osteoarthritis with overweight, race, and occupation, all of which have been suggested by smaller cross-sectional studies. They then focused specifically on those factors. For overweight, they found a strong association between current obesity and osteoarthritis of the knee, with a dose-response effect not previously assessed. This association was also seen for self-reported minimum adult weight, a proxy for long-term obesity, and was present in persons with asymptomatic osteoarthritis of the knee. These findings strongly suggest that obesity is causative. HANES I was the first study in which racial differences in osteoarthritis of the knee could be assessed within the same country. The black women who were studied had an increased risk of disease (odds ratio (OR) = 2.12, 95% confidence interval (CI) = 1.39-3.23) after controlling for age and weight, although the black men did not. Finally, the authors used the US Department of Labor Dictionary of Occupational Titles to obtain characterizations of the physical demands and knee-bending stress associated with occupations and to study the relation between physical demands of jobs and osteoarthritis of the knee. They found for persons aged 55-64 years an association between knee-bending demands and osteoarthritis of the knee (men, OR = 2.45, 95% CI = 1.21-4.97; women, OR = 3.49, 95% CI = 1.22-10.52). Since such occupational physical demands are common, the authors conclude that they may be associated with a substantial proportion of osteoarthritis of the knee.