Background: NSABP B-41, randomly assigned 529 patients with HER2 positive breast cancer to receive neoadjuvant trastuzumab, lapatinib, or the combination, with weekly paclitaxel following doxorubicin and cyclophosphamide x 4. No significant difference in pCR was found among three arms (Robidoux 2013), but overall survival was significantly increased for patients who obtained a pCR (Robidoux 2016). This study evaluated outcomes in B-41 based on intrinsic subtype. Methods: H&E slides were reviewed by study pathologist for areas of invasive carcinoma having > 10% tumor cells. RNA was isolated and hybridized to the “RUO-PAM50” CodeSet (NanoString Technologies). Intrinsic PAM50 subtype was determined for each sample. Contingence-table analyses were used to compare pCR breast and nodes (ypT0/is ypN0) among intrinsic subtypes and Kaplan-Meier estimates and Cox models were used to compare event-free survival and overall survival among subtypes. Results: Core biopsy samples from 276 patients, prior to therapy, were evaluated. Intrinsic subtype was determined in 271: 197 (73%) were HER2 enriched (E) (97 ER negative and 100 ER positive), 26 (10%) basal-like, 23 (8%) luminal A, and 25 (9%) luminal B. pCR was 61% in HER2-E (69% in ER negative and 53% in ER positive) and 26% in others (basal-like, luminal A/B) (p < 0.001). In HER2-E, patients on trastuzumab-based arms had higher pCR than those on the lapatinib arm (67% vs 49%, p = 0.02). Patients with basal-like tumors had higher pCR than those with luminal A/B tumors (38.5% vs 13% & 24%) but their long-term prognosis is poor, compared with other three subtypes: five-year EFS = 0.724 [0.506, 0.858] (p = 0.28) and five-year OS = 0.808 [0.598, 0.915] (p = 0.005). In HER2-E, EFS and OS were no different between the three arms. Conclusions: HER2-E subtype is a good marker for predicting benefit from HER2 targeted therapies, particularly if trastuzumab-based. Whether tumors with other subtypes are suitable for HER2 targeted therapies needs further investigation. Survival outcomes are limited by the small number of events and would benefit from meta-analysis with other similar studies. Clinical trial information: NCT00486668