Infiltrative astrocytic neoplasms are the most common malignancies of the central nervous system. They remain clinically problematic because of their involvement of brain structures critical to proper cognitive, behavioral, and motor function; their widely invasive properties, which make them difficult to resect totally; and their nearly inevitable biologic progression in spite of adjuvant therapy. Glioblastoma multiforme (GBM, World Health Organization grade IV), the most malignant form of infiltrating astrocytoma, can present as a high-grade lesion from the outset (so-called de novo GBM) or can evolve from a lower grade precursor lesion (secondary GBM). Molecular genetic investigations suggest that GBM is best regarded as a clinicopathologic entity composed of multiple molecular genetic subsets. Molecular alterations associated with progression to GBM and that define genetic subsets include epidermal growth factor receptor amplifications, p53 mutations, retinoblastoma pathway alterations [most commonly, p16(CDKN2A) losses], and chromosome 10 alterations, including PTEN mutations. Despite the wide range of genetic events that ultimately lead to GBM, the vascular changes that evolve are remarkably similar. Microvascular hyperplasia is spatially and temporally associated with pseudopalisading necrosis in GBM and is believed to be driven by hypoxia-induced expression of proangiogenic cytokines such vascular endothelial growth factor. In addition, genetic alterations in GBM are thought to contribute directly or indirectly to angiogenic dysregulation. Both p53 mutations and genetic losses on chromosome 10 may tip the balance toward an angiogenic phenotype through upregulation of proangiogenic factors and/or downregulation of angiogenesis inhibitors. Understanding genetic events and their relation to angiogenic regulation in astrocytic neoplasms may eventually lead to therapies that are specifically directed at molecularly defined subsets of these diseases.