Assessing the presence of circulating tumour DNA in plasma samples enables the early prediction of relapse among patients who received therapy for breast cancer and are apparently disease-free. These findings stem from a study involving 55 women who had been treated with neoadjuvant chemotherapy and surgery for early breast cancer. From an analysis of the mutations present in the primary tumours of each patient, the researchers developed personalized digital PCR assays that they later used to monitor the existence of minimal residual disease in plasma samples collected either at a single time point after surgery or serially during follow-up. This approach led to detection of metastatic relapse with high accuracy and sensitivity, particularly when performed in serial samples (for which detection occurred at a median of 7.9 months before clinical relapse). This procedure could be used to identify patients at high risk of relapse and to tailor therapy to their specific mutation profiles.