Treatment advances have differed by breast cancer subtype. We examined the impact of race on survival of women with metastatic breast cancer by disease subgroup. Using surveillance, epidemiology, and end results -Medicare data, we included white and black patients aged ≥66 with de novo metastatic breast cancer diagnosed between 1998 and 2009. Using trastuzumab as a proxy for human epidermal growth factor receptor 2 (HER2)-positive status, we defined three disease subgroups: (1) HER2-positive (received trastuzumab), (2) HER2-negative/unknown (never received trastuzumab)/hormone receptor (HR)-positive, and (3) HER2-negative/unknown/HR-negative. Multivariate Cox proportional models assessed the impact of race on overall survival (OS) and breast cancer-specific survival by subgroup. We also examined trastuzumab treatment patterns. Among 4364 women (86 % white, 14 % black), 9 % had HER2-positive, 72 % had HER2-negative/unknown/HR-positive, and 18 % had HER2-negative/unknown/HR-negative tumors. Patients with HER2-positive disease experienced longer median OS compared with others: 2.4 versus 1.8 years for women with HER2-negative/unknown/HR-positive and 0.5 years for women with HER2-negative/unknown/HR-negative disease (P < 0.001). Racial differences in OS were only observed among patients with HER2-positive tumors: median OS: 1.4 versus 2.7 years for black and white women, adjusted hazard ratio 1.45; 95 % Confidence interval 1.01-2.08. Results for breast cancer-specific survival were similar. We also observed racial differences in trastuzumab utilization, with longer median time to trastuzumab initiation and lower likelihood of continuation over time for black (vs. white) patients. Among women with de novo metastatic breast cancer, racial differences in survival were only apparent for those with inferred HER2-positive tumors. Further study of how treatment patterns affect outcomes is warranted.