Abstract PURPOSE: Whether patients with acute myocardial infarction who are treated by cardiologists have better outcomes than patients treated by generalist physicians is controversial. Because some of the survival benefit associated with cardiology care may be due to baseline differences in patient characteristics, we evaluated how differences in case-mix of comorbid illness and functional limitations may explain the association between specialty care and survival. MATERIALS AND METHODS: We examined the records of 109,243 Medicare beneficiaries hospitalized for myocardial infarction from 1994 to 1995 from the national Cooperative Cardiovascular Project to evaluate the association of physician specialty with 30-day and 1-year mortality. We assessed the extent to which this relation was mediated by differences in the use of guideline-supported therapies (aspirin, beta-blockers, reperfusion, angiotensin-converting enzyme inhibitors) or differences in the clinical characteristics of the patients. RESULTS: Patients who had board-certified cardiologists as attending physicians had the least number of comorbid conditions, whereas patients who had general practitioners or internal medicine subspecialists as attending physicians usually had the most comorbidities. Cardiologists had the greatest use of most guideline-supported therapies, and general practitioners had the lowest use. After adjustment for severity of myocardial infarction, clinical presentation, and hospital characteristics, patients treated by cardiologists were less likely to die within 1 year (relative risk [RR] = 0.92, 95%, confidence interval [CI]: 0.89 to 0.95), and patients cared for by other general practitioners were more likely to die within 1 year (RR = 1.09, 95% CI: 1.03 to 1.14), than patients cared for by general internists. After adjusting for additional measures of comorbid illness and functional limitations, the 1-year survival benefit associated with cardiology care was attenuated relative to internists (RR = 0.97, 95% CI: 0.94 to 1.0), and the excess mortality associated with general practitioners decreased (RR = 1.05, 95% CI: 1.00 to 1.11). After further adjustment for the use of guideline-supported therapies, both differences in 1-year survival between patients treated by cardiologists or general practitioners were not significantly different from those of patients treated by internists. CONCLUSION: Studies comparing outcomes by physician specialties that do not adjust adequately for differences in patient characteristics may attribute more benefit than is appropriate to specialists who treat patients who have fewer comorbid conditions. Some of the remaining benefit—at least among patients with myocardial infarction—may be attributable to greater use of recommended therapies.