Continuing evaluations of antiviral agents for treatment of herpes simplex encephalitis (HSE) provided an opportunity to collect clinical data from 113 patients in whom the diagnosis was proved by viral isolation. Occurrence of HSE was in all ages and in both sexes and was nonseasonal. Characteristically, patients had behavioral changes, fever, confusion, speech disturbances, and, less frequently, seizures. The EEG was the most useful neurodiagnostic aid followed by technetium and computed axial tomographic scans. Employing a logistic regression model for variable selection, the diagnosis could be predicted by clinical findings and neurodiagnostic tests in 83% of the proved cases, but the evidence in 25% was falsely positive. There was evidence of localization by either clinical or neurodiagnostic assessment in all patients with proved disease. Among patients wtih negative findings for HSE, similar focal findings predominated in all but a few. The CSF and brain scans were normal in many patients with proved HSE. This extensive clinical experience in patients wtih diagnosis proved by viral isolation shows that diagnosis cna be confirmed only by brain biopsy.