CONTEXT Perineural invasion and vascular invasion may be adverse prognostic factors in patients with oral cavity squamous cell carcinoma. However, the incidence of perineural and vascular invasion varies in the literature, and the use of immunohistochemistry to enhance their detection has not been evaluated in oral cavity squamous cell carcinomas. OBJECTIVE To determine if the previously assessed incidence of perineural and vascular invasion in cases of oral cavity squamous cell carcinoma would be increased by re-review of the original routinely hematoxylin-eosin-stained sections as well as review of slides stained immunohistochemically with S100 and CD31 to enhance visualization of nerves and vessels. DESIGN Forty cases of oral cavity squamous cell carcinoma in which the status of perineural and vascular invasion had been part of the original pathology report were reviewed. All original routinely stained slides were reviewed as well as S100- and CD31-stained sections of each case's tissue blocks that contained tumor. RESULTS Perineural invasion was identified in 30% (12/ 40) of tumors in the original reports, 62% (25/40) of the authors' re-review of the same slides, and 82% (33/40) when cases were stained with S100. Vascular invasion was identified in 30% (12/40) of tumors in the original reports, 35% (14/40) of the authors' re-review of the same slides, and 42% (17/40) when cases were stained with CD31. False-positive and false-negative results were common in the original reports. The number of foci of both types of invasion was related to its discovery in the original reports. Vascular invasion, but not perineural invasion, was significantly associated with death at 5-year follow-up. CONCLUSIONS Although careful re-review of routinely stained slides will detect a significant number of cases of perineural and vascular invasion, immunohistochemical enhancement further improves the accuracy of the determination.