Population based mammographic screening has seen an increased detection of ductal carcinoma in situ (DCIS) and known or putative pre-invasive lesions such as atypical ductal hyperplasia, lobular hyperplasia and papillomas. Importantly, screening has also led to over-diagnosis of lesions that would never have become symptomatic within the lifespan of the individual and results in significant unnecessary interventions. However, these lesions cannot be ignored since many confer an increased risk of invasive carcinoma and this makes clinical management problematic. For example, currently, only histopathology guides the decision to excise non-malignant lesions, and lesion classification and the decision making process has been shown to vary greatly. Therefore, having objective markers of increased risk of progression would enable more aggressive treatment where merited, while reducing unnecessary procedures. Reliable discrimination of high risk lesions and appropriate treatment should lead to reduced incidence of malignant breast cancer and associated mortality. Patterns of somatic genomic aberrations can provide a reliable reflection of the underlying biology of neoplasms. We are therefore conducting high-resolution genome wide copy number analyses of these lesions to identify genomic aberrations that are associated with progression to invasive disease. For example, we have shown that DCIS cases where there was a recurrent cancer had more numerous copy number events than the DCIS without a recurrence. Furthermore these copy number changes appeared to be non-random with several genomic regions showing an increase in frequency in recurrent cases.