The authors examine three hypotheses regarding race differences in utilization of coronary angiography (CA): (1) patients with a cardiology consultation are more likely to obtain a referral for CA, (2) African American patients are less likely to have a cardiology consultation, and (3) among patients referred for CA, there is no difference by race in receipt of the procedure. To determine if they obtained a referral for or received CA, 2.623 candidates for CA were followed. Multivariate models were estimated using logistic regression. Cardiology consultation was associated with referral for CA (OR = 5.1, p < .001). White patients had higher odds of cardiology consultation (OR = 2.2, p < .001). The racial disparity was reduced among patients who received a referral (OR = 1.4, p < .05). Researchers must eliminate racial differences in access to specialty care and variation in referral patterns by physician specialty, and efforts must be targeted to those specialties where greater disparities exist.