To a large extent cancer conforms to evolutionary rules defined by the rates at which clones mutate, adapt and grow. Compared to species evolution however, cancer is a particular case, due to the vast population size, chromosomal instability, and the potential for phenotypic plasticity. Nevertheless, an evolutionary framework is a powerful aid in our understanding of cancer progression and therapy failure, and could be applied to predict individual tumour behaviour and aid treatment strategies. Introduction Tumours are composed of subpopulations of cells (subclones) that may be distinguished by a variety of features impacting their phenotype, including genetic alterations. Genetic intratumour heterogeneity (ITH) has been documented across most cancers (reviewed in 1) and acts as a substrate for clonal evolution. The fundamental biological mechanisms underlying clonal evolution [G] in cancer are similar to those that underpin the evolution of asexually-reproducing species: replication, heritable variation, genetic drift [G], selection [G] and environmental changes. Central to the neo-Darwinian synthesis of evolutionary biology is the paradigm of molecular evolution [G], which links Mendelian genetics to Darwinian adaptation. Molecular evolution is relevant to cancer because the use of genomic sequencing is a key technology to understand temporal and spatial patterns of somatic evolution [G]. At the core of molecular evolution, in turn, is theoretical population genetics, which has been the fundamental mathematical formalism to describe evolution for the past 90 years 2,3 The same theoretical framework has been used to understand clonal evolution in cancer 4 5 6 7 8 9 10 11. The study of the evolutionary dynamics of cancer clones is fundamentally concerned with the relative frequencies of cancer subpopulations over space