This new method for risk stratifying patients could lead to more personalised patient care. This is the first study to analyse all the relevant changes in DNA across the entire cancer genome, rather than targeted regions, to classify patients with cancer and link these subgroups to clinical outcomes. They identified 186 distinct and recurrent genomic alterations and used these to define five genomic subgroups of CLL that associate with different clinical outcomes. This study paves the way for routine clinical application of whole genome analyses for risk stratification in other cancer types. It will also aid in selecting the most appropriate patients for trials of new, targeted therapies.