Background. Hip and knee joint replacement rates vary by demographic group. This article describes the epidemiology of need for joint replacement, and of subsequent receipt of a joint replacement by those in need. Methods. Data from the Health and Retirement Study were used to assess need for hip or knee joint replacement in a total of 14,807 adults aged 60 years or older in 1998, 2000, and 2002 and receipt of needed surgery 2 years later. “Need” classification was based on difficulty walking, joint pain, stiffness, or swelling and receipt of treatment for arthritis, without contraindications to surgery. Results. Need in 2002 was greater in participants who were older than 74 years (vs 60–64: adjusted odds ratio 2.06; 95% confidence interval, 1.68–2.53), women (vs men: 1.81; 1.53–2.14), less educated (vs college educated: 1.27; 1.06–1.52), in the poorest third (vs richest: 2.20; 1.78–2.72), or obese (vs nonobese: 2.39; 2.02–2.81). One hundred sixty-eight participants in need received a joint replacement, with lower receipt in black or African American participants (vs white: 0.47; 0.26–0.83) or less educated (vs college educated: 0.65; 0.44–0.96). These differences were not explained by current employment, access to medical care, family responsibilities, disability, living alone, comorbidity, or exclusion of those younger than Medicare eligibility age. Conclusions. After taking variations in need into consideration, being black or African American or lacking a college education appears to be a barrier to receiving surgery, whereas age, sex, relative poverty, and obesity do not. These disparities maintain disproportionately high levels of pain and disability in disadvantaged groups.