Gene amplification occurs in 45-50% of malignant human gliomas (MHG). In the present study, 64 genetically characterized gliomas were evaluated to determine if tumors with amplification of the epidermal growth factor receptor (EGFR), N-myc, c-myc, or gli genes had distinctive histopathologic features. There was no significant difference in age (p = 0.10) or gender (p = 0.78) between patients whose tumors contained amplified genes and those whose tumors did not exhibit this characteristic. Although the patients with amplified genes in their tumors survived slightly longer than patients whose tumors had no detectable gene amplification, these differences were not statistically significant (p = 0.21). The 28 tumors with amplification included 24/48 (50%) glioblastoma multiforme, 2/6 (33%) anaplastic astrocytomas and 2/5 (40%) gliosarcomas. No amplification was seen in one oligodendroglioma, three anaplastic mixed gliomas or one giant cell glioblastoma multiforme. Necrosis and endothelial proliferation were equally prevalent among tumors with and without amplification. Comparison of tumors with gene amplification and tumors without this characteristic revealed similar distributions of most morphologic cells types. Although prominent perivascular lymphocytic infiltrates were more frequent in tumors without amplification, this association was of borderline significance statistically. In situ hybridization of tumors with amplification using an EGFR mRNA probe showed intense labeling of different neoplastic cell types including fibrillary and protoplasmic astrocytes, gemistocytes, anaplastic cells, and multinucleated giant cells. Non-neoplastic cells such as hyperplastic endothelium within the tumors did not express detectable EGFR mRNA. These studies demonstrate that (a) cells with quite different morphology within the same tumor can contain the same genetic alteration; (b) tumors of identical histological appearance may have arisen and evolved by different molecular mechanisms; and (c) molecular analyses are necessary to evaluate gene amplification in MHG since this characteristic cannot be accurately predicted by the morphologic or clinical criteria used in this study.