Studies have shown that ALDH1A1 expression in the breast is associated with worse clinical outcome. ALDH1A1 inactivates cyclophosphamide, which is an integral agent in breast cancer chemotherapy regimens. The purposes of this study were to verify these results, to correlate ALDH1A1 expression with clinical outcome in patients treated with cyclophosphamide as part of the chemotherapy (adjuvant or neoadjuvant), and to evaluate ALDH1A1 as a useful marker to predict the clinical outcome of breast cancer subsets. A total of 513 primary breast cancers were studied. Tissue microarrays of the studied cases were stained with ALDH1A1. Key clinicopathological information was obtained. Disease-free survival and overall survival were calculated. Patients with neoadjuvant therapy who had substantial residual cancer burden (RCB) were included in the study. Fisher's exact test and Kaplan-Meier methods were used for statistical analysis. ALDH1A1 was expressed in 53 (10%) patients, with a higher frequency in triple negative, followed by HER2+, and finally hormonal receptor+/HER2- (P<0.0001). Tumors with advanced stage, node-positive, or larger tumor size were correlated with ALDH1A1 expression (P=0.006, P<0.0001, and P=0.05, respectively). ALDH1A1 expression was also correlated with worse disease-free survival (P<0.006) and overall survival (P<0.01) in patients who were treated with neoadjuvant chemotherapy. In all, 8 of 22 (36%) received neoadjuvant chemotherapy and died of disease-expressed ALDH1A1 (P=0.008). Similarly, 8 of 23 (35%) who received neoadjuvant chemotherapy and had tumor recurrence expressed this marker (P=0.002). The risk of recurrence was fivefold greater than negative ALDH1A1 tumors. The risk of recurrence became 11-fold greater when cyclophosphamide but not trastuzumab was part of the regimen. Our results are consistent with previous studies. Moreover, we found that ALDH1A1 could be a useful marker to predict worse clinical outcome after chemotherapy in the neoadjuvant setting with substantial RCB. However, a larger cohort is required to verify our results.