Triple-negative breast tumours are defined as oestrogen receptor (ER)-negative, progesterone receptor (PgR)-negative and human epidermal growth factor receptor 2 (HER2)-negative. Beyond this common negative triplet, however, triple-negative breast cancers (TNBC) are not an homogeneous subgroup. There is substantial intra-subgroup diversity in tumour biology, prognosis and treatment sensitivity. Triple-negative breast cancer accounts for 15−20% of all breast cancer cases; however, it accounts for a disproportionate number of breast cancer relapses and deaths. Studies in early breast cancer have shown patients with TNBC to have a worse outcome compared with patients with non-TNBC [1]. Interestingly, neo-adjuvant studies highlight a subgroup of TNBC patients, approximately 20%, with highly chemosensitive disease and post-chemotherapy outcomes as favourable as their non-TNBC counterparts [2]. A key issue with TNBC is the accurate diagnosis of histological subtype and ER, PgR and HER2 status. ER, PgR and HER2 measurement has been associated with concerns regarding reliability and reproducibility, with remarkable reported rates of discordance. Furthermore, prior arbitrary thresholds for defining positivity for ER and PgR status, including 1%, 10% and 20%, have varied between laboratories and clinical trials. Clarity is offered by recent St Gallen and ASCO-CAP guidelines that define endocrine responsiveness as 1%. TNBC includes several morphological subtypes. Most are high-grade ductal carcinomas (not otherwise specified) but the triple-negative phenotype also occurs in medullary, apocrine and squamous cell carcinomas. Some subtypes, such as medullary carcinoma, are associated with a better prognosis and may not require aggressive chemotherapy. In contrast to endocrine-sensitive and HER2-positive breast cancers, which are defined by the presence of specific targets for matched targeted therapy, there is currently no targeted therapy for TNBC. There are promising new targeted biological therapies, such as PARP inhibitors; however, chemotherapy remains the mainstay of systemic therapy for TNBC. The biological diversity displayed within TNBC does not support such an empirical approach; however, a lack of robust predictive biomarkers prevents a more individualised approach to the choice of adjuvant therapy. Predictive markers could guide chemotherapy use by identification of the subset of TNBC patients with chemosensitive disease, or more specifically to predict sensitivity or resistance to individual chemotherapeutic agents. However, currently there are no clinical tools to recommend use, or omission, of specific chemotherapeutics. Conventional thirdgeneration combination chemotherapy is the standard treatment. Rates of pathological complete response to conventional neoadjuvant anthracycline/taxane-based chemotherapy are reported to be within the range of 13−45% [2]. Optimal adjuvant systemic therapy for TNBC is difficult to define because of a lack of robust prospective data in triple-negative restricted trial populations. Most clinical data derive from retrospective exploratory subgroup analyses from biologically unselected trial populations, and small underpowered studies. Most data reveal a highly chemosensitive TNBC subset. Lack of randomised phase III trials with a standard therapy control arm make it difficult to determine whether this increased sensitivity is agent-specific or whether it reflects general sensitivity to chemotherapy (Fig. 1). Triple-negative breast cancers are associated with a high proliferative rate, which may render them particularly sensitive to chemotherapy. Furthermore, a subset of TNBCs displays phenotypic overlap with BRCA1-mutation-related breast cancers. BRCA1 is critical in homologous recombination, a cellular process of double-strand DNA damage repair, and in the cell cycle arrest necessary for DNA repair.