Data from the National Health and Nutrition Examination Survey I, 1971-1975 (NHANES I) were used to analyze associations among highest education level and arthritis. The dependent variables indicated whether the respondent had ever been diagnosed with any form of arthritis by a physician (10,678 women and 7243 men) or whether physician X-ray readings suggested arthritis of the knee (3491 women and 3119 men). These variables did not distinguish between osteo- and rheumatoid arthritis. It is likely that the great majority of the sample reporting or diagnosed with arthritis had osteoarthritis. There were strong univariate correlations between answers to the general arthritis question and the knee question on the one hand and gender, age, body mass, schooling, income and employment on the other. Respondents' education level was found to be strongly and negatively associated with self-reported arthritis in the larger samples both before and after controls were entered for employment, income and potential biological risk factors. The association between self-reported arthritis or arthritis of the knees and education was weaker for men, but not for women after employment and income were accounted for. When body mass was accounted for, the association between self-reported arthritis or arthritis of the knees and education was weaker among women but not men. Long-run preventive strategies to combat osteoarthritis ought to consider investments in education.