Background: According to gene expression profiling, HER2+ BC is heterogeneous and appears to diverge by HR status. Methods: We evaluated 3394 pts who presented to NCCN centers with stage I-III HER2+ BC between 2000-07. Pts were classified as HR+ (ER+ and/or PR+) and HR- (ER- and PR-). Chi-square, univariate logistic regression, log-rank test, and Cox hazard proportional regression were used for analysis. Results: Median follow-up was 51 months. 59% of patients had HR+ and 41% HR- disease respectively. Pts with HR- BC were more likely to be postmenopausal and to present with higher stage and high grade disease (p<0.001). Most pts received adjuvant or neoadjuvant therapy; 44% received adjuvant trastuzumab. Recurrences were recorded for 458 pts. HR- patients were more likely to recur first in the central nervous system (CNS) (OR: 1.8, 95% CI: 1.1, 2.9; p= 0.03) and less likely to recur in bone (OR: 0.5, 95% CI: 0.3, 0.8; p<0.01). No differences in risk of lung or liver recurrence were observed. Combining first and subsequent sites of recurrence, the difference in CNS involvement was lost (p=0.107) but HR- were more likely to experience lung involvement (OR: 1.5, 95% CI: 1.0, 2.2; p= 0.05). After adjusting for age, year of diagnosis (y), race, stage, and grade, HR- had worse survival after initial BC diagnosis than HR+ pts (Hazard Ratio of death [HRd] 1.4, 95% CI: 1.14, 1.7; p<0.01). However, the risk of death was not proportional over time with HR- having significantly increased hazard in the first five years: HRd 0-2 y 1.9 [1.3, 2.9]; p< 0.01; HRd 2-5y 1.5, [1.2, 2.0]; p<0.01; HRd 5+ y, 0.8, [ 0.6, 1.2], p=0.29. Conclusions: Clinicopathological features, sites of recurrence, and risk of death over time for HER2+ BC differed by HR status. This suggests that HR status in HER2+ BC is clinically relevant. These differences should be further explored from a mechanistic and therapeutic standpoint.