Glioblastomas (GBM) are the commonest and most aggressive primary tumour of the brain. They are associated with an appalling prognosis with only survivals of a matter of weeks in untreated patients, improving to 14 - 16 months with radiotherapy with concomitant and adjuvant chemotherapy. In this era of molecular biology diagnosis still relies on the WHO Classification system that only requires a tumour to have features of an astrocytic tumour with either necrosis or microvascular proliferation on light microscopy. Yet it is clear that GBMs form a very heterogenous group of tumours. Although there are recognized histological variants (e.g. giant cell GBM, GBM with oligodendroglial differentiation) it is doubtful how useful these are for predicting outcome. It is clear we need to be able to better subtype GBMs to consider individualizing treatment. We have known for some time that glioblastomas can arise from two pathways – either as a primary ‘de novo’ GBM or a ‘secondary’ GBM from a pre-existing low grade glioma. These two subtypes have different mutations and activate different pathways within the cells. The discovery of mutations of the isocitrate dehydrogenase (IDH) gene, an early marker of astrocytic tumor development, has lead to the realization that some GBMs have this mutation. This cohort of patients have a far better prognosis with a median survival of 31 months compared to the 15 months in IDH-wild type.