The great heterogeneity of breast cancer makes it impossible to firmly predict which patients with early-stage tumours will or will not need systemic treatments according to the conventional prognostic factors currently employed. In fact, a substantial percentage of patients receive medical treatment for a disease that will not relapse, while another proportion of patients regarded as having good prognostic factors according to the classic criteria do not receive treatment and suffer disease relapse. Considering that most oncological treatments have short- and long-term toxic effects, new methods capable of offering a more precise prognosis need to be developed. The individualisation of the diagnosis of patients with breast cancer based on molecular and gene expression studies is bringing about a veritable revolution in our understanding of the biology of the disease. The new molecular classification of breast cancer, based on these profiles, allows us to establish a prognosis according to the genetic characteristics of each tumour. Such individualisation of the diagnosis of patients with breast cancer will lead to the application of more specific treatments, thereby improving patient survival with lesser toxicity and increased economic savings. Of the different genetic analytical tests available, MammaPrint has been shown to be the option offering the most information on the behaviour of early breast cancer; as a result, it is the most useful technique in deciding the need for oncological treatment as a complement to surgery.