The response to neoadjuvant chemotherapy (NAC) is highly correlated with survival in breast cancer (BC) patients. The early prediction of the response to NAC could facilitate treatment adjustments on a patient-by-patient basis, which would improve patient outcomes and survival. Conventional techniques used for detecting circulating microRNAs (miRNAs), which act as biomarkers for the early prediction of NAC efficacy in BC patients, are associated with limitations such as low sensitivity and specificity. We designed a highly sensitive graphene oxide (GO) based qRT-PCR method for detecting miRNAs associated with the chemotherapeutic response in BC patients. The results showed that miRNA levels at both the baseline and end of the first NAC cycle could help distinguish NAC responders from NAC nonresponders; BC patients with lower plasma miRNA levels were more likely to achieve pathological complete remission. Thus, GO-based qRTPCR could facilitate early prediction of NAC efficacy in BC patients.