While most authors currently classify dural-based hemangiopericytoma (HPC) as a distinct entity rather than as a subtype of meningioma, the histogenesis of HPC has long been debated. We have recently shown that meningiomas contain frequent mutations of the neurofibromatosis 2 gene, while HPCs do not, suggesting that HPC is genetically distinct from meningioma. In the present study, we evaluated a series of 31 dural HPCs (including 3 pairs of primary and recurrent tumor) and 26 meningiomas for alterations in the cell-cycle regulatory genes CDKN2/p16 and p53. Homozygous deletions of the CDKN2/p16 gene were detected using a comparative multiplex polymerase chain reaction assay in 7 of 28 primary HPCs (25%), but in only one of 26 meningiomas (P = 0.03). Among the HPCs with recurrence, 1 pair of 3 had a homozygous CDKN2/p16 deletion. The 1 meningioma with a CDKN2/p16 deletion was a meningothelial meningioma, without atypical or malignant features. Single-strand conformational polymorphism analysis of all three exons of CDKN2/p16 and exons 5-8 of p53 revealed no mutations in either HPCs or meningiomas. These results illustrate that homozygous deletions of CDKN2/p16 occur in HPCs and suggest that alterations of the p16-mediated cell-cycle regulatory pathway may underlie the formation or progression of some HPCs. The data also provide further genetic evidence that HPC is not a subtype of meningioma.