Therapeutic decision for adjuvant systemic therapy for breast cancer involves assessment of baseline risk and estimated benefit from systemic therapy. Molecular profiling studies have clearly demonstrated heterogeneity of chemotherapy response across different molecular subtypes of breast cancer. Meta-analyses of publicly available data from gene expression profiling studies have demonstrated that breast cancer can be divided into 4 basic categories based on expression levels of estrogen receptor (ER), HER2, and proliferation-associated genes; ER-, HER2+, ER+/HER2-/low proliferation, and ER+/HER2-/high proliferation. Notably ER- or HER2+ tumors are associated with high levels expression of proliferation genes, although there is a wide spectrum of expression levels of proliferation genes among ER+/HER2- tumors. Estrogen receptor-positive/HER2-/low-proliferation tumors are associated with a favorable prognosis. Synthetic lethal screening approach has demonstrated that most of the chemotherapeutic agents do not have specific molecular targets. Therefore, it could be hypothesized that chemosensitivity would be largely dictated by proliferation activity of tumor cells. Therefore, tumors with ER-, HER2+, or ER+/HER2-/high proliferation gene expression profile can be categorized as chemosensitive tumors, whereas ER+/HER-/low proliferation tumors categorized as chemoresistant. Therefore, clinical utility of gene expression profiling is mainly in aiding the chemotherapy decision for ER+ patients. Although evidence from prospective randomized clinical trials are lacking, because of the excellent baseline prognosis of patients with ER+/HER2-/low proliferation tumors when treated with endocrine therapy and because of scientific evidence of chemoresistance of these tumors, a comfort zone has been reached among oncologists to allow clinical use of gene expression tests to identify patients who do not require chemotherapy among node-negative ER+ patients. However, these tools are still probabilistic at best in their performances, and one cannot exactly predict which patient will have recurrence after assigned therapies until the time of recurrence. Therefore, strategies have to be established to identify patients who will fail standard chemoendocrine therapies among high-risk patients (ER+/HER2-/high proliferation, HER2+, or ER-) before recurrence events. Neoadjvant therapy can provide such venue because regardless of regimens used the prognosis of those achieving complete pathological response is excellent. Postneoadjuvant setting can be then used for patients with gross residual disease to test novel therapeutic agents.