Background Although there is a wide literature demonstrating sex and race differences in the receipt of invasive cardiac tests and treatments, much less is known about the influence of such characteristics on receipt of a stress test, the first event in the diagnostic/treatment cascade for many patients. We explored the influence of patient characteristics on receipt of a stress test, with special attention to sex and race. Methods We performed a nested case-control study of Medicare beneficiaries who were aged 66 years and older during 1999-2001 and were free of cardiac diagnoses and procedures for at least 1 year. Cases were recipients of a stress test. Results Cases were younger, less likely to be female or black, but more likely to live in high-income, highly educated, and urban areas than controls. Nonblack men were more likely to receive a stress test than women and black men, controlling for age, area characteristics, and clinical characteristics (odds ratio for nonblack men compared with black women 1.71). These results were not explained by physician visit frequency. Conclusions Efforts at minimizing disparities in cardiac care must attend to what is, for many patients, the entry into the cardiac care system: the stress test. Our findings suggest that simple “access,” as measured by physician visit frequency, is not a rate-limiting factor.