Breast cancer is one of the leading causes of mortality in women worldwide, neoadjuvant chemotherapy has emerged as an option for the management of locally advanced breast cancer. Extensive efforts have been made to identify new molecular markers to predict the response to neoadjuvant chemotherapy. Transcripts that do not encode proteins, termed lncRNAs, have been shown to display abnormal expression profiles in different types of cancer but their role as biomarkers in response to neoadjuvant chemotherapy has not been extensively studied. Herein, lncRNA expression was profiled using RNA-Seq in biopsies from patients who subsequently showed either response or no response to treatment. The GATA3-AS1 transcript was overexpressed in the nonresponder group and was the most stable feature when performing selection in multiple random forest models. GATA3-AS1 was experimentally validated by RT-qPCR in an extended group of 68 patients. Expression analysis confirmed that GATA3-AS1 is overexpressed primarily in patients who were nonresponsive to neoadjuvant chemotherapy with a sensitivity of 92.9%, specificity of 75.0% and an AUC of ∼0.90 as measured by ROC curve analysis. The statistical model based on Luminal B-like patients adjusted by menopausal status and phenotype (OR= 37.49; 95% CI= 6.74-208.42 P= 0.001), GATA3-AS1 was established as an independent predictor of response. Thus, lncRNA GATA3-AS1 is proposed as a potential predictive biomarker of nonresponse to neoadjuvant chemotherapy.