The refinement of the classification, the risk of relapse and the prediction of response to multidisciplinary treatment for early breast cancer has been the major theme of the 14th St Gallen International Breast Cancer Consensus Conference 2015. The meeting, held in Vienna, assembled 3500–4000 participants from 134 countries worldwide. It culminated, on the final day, with the International Consensus Session, delivered by 40–50 of the world’s most experienced opinion leaders in the field of breast cancer treatment. The panelists addressed the “semantic” classification of breast cancer subtypes by pathology-based biomarkers (e.g. estrogen receptor, progesterone receptor and HER2) vs genomic classifiers. They also refined the biomarker prognostication dissecting the impact of the various gene signatures and pathologic variables in predicting the outcome of patients with early breast cancer in terms of early and late relapse. Finally they addressed the challenges stemming from the intra- and inter-observer variability in the assessment of pathologic variables and the role of gene signatures for the prediction of response to specific therapeutic approach such as endocrine therapy and chemotherapy and for personalizing local treatment of patients with early breast cancer. The vast majority of the questions asked during the  onsensus were about controversial issues. The opinion of the panel members has been used to implement guidance for treatment choice. This is the unique feature of the St. Gallen Consensus, ensuring that the resulting recommendations will take due cognizance of the variable resource limitations in different countries. Information derived from evidence based medicine and large meta-analyses is of obvious and enormous value. The weakness of this approach is that it gives particular weight to older trials (which have accumulated more event endpoints) and is frequently unable to collect sufficient detail on the patients and tumors in the trials to allow assessment of whether the treatments which are better on average offer equal value to all currently definable patient subgroups. What St Gallen can provide is clinically useful updated breast cancer treatment consensus for the majority of patients treated outside of clinical trials (>90%) in most countries.