BACKGROUND: Knowledge of the epidemiology and clinical characteristics of varicella zoster virus (VZV) encephalitis remains limited. METHODS: Nationwide prospective cohort study of adults treated for microbiologically confirmed VZV encephalitis at Danish departments of infectious diseases from 2015 to 2019. Modified Poisson regression analysis was used to compute adjusted relative risks (RRs) of unfavorable outcome. RESULTS: We identified 92 adults (49% female) with VZV encephalitis, yielding an incidence of 5.3/1 000 000 per year (95% CI, 4.2-6.6). Median age was 75 years (IQR, 67-83) and immunocompromising conditions were frequent (39%). Predominant symptoms were confusion (76%), headache (56%), nausea (45%), gait disturbance (42%), and personality changes (41%). Cranial imaging showed cerebral vasculitis (including infarction and hemorrhage) in 14 (16%) patients and encephalitic abnormalities in 11 (13%) with predilection for the brainstem and deep brain structures. Intravenous acyclovir treatment was initiated a median (IQR) of 13.4 hours (5.2-46.3) since admission, while cranial imaging and lumbar puncture were performed after 6.3 hours (2.5-31.0) and 18.5 hours (4.9-42.0). In-hospital, 1-month, and 3-month mortalities were 4%, 9%, and 11%, respectively. Unfavorable outcome (Glasgow Outcome Score of 1-4) was found in 69% at discharge, with age (adjusted RR [aRR], 1.02; 95% CI, 1.01-1.03), vasculitis (aRR, 1.38; 95% CI, 1.02-1.86), and Glasgow Coma Scale (GCS) <15 (aRR, 1.32; 95% CI, 1.01-1.73) identified as independent risk factors. CONCLUSIONS: VZV encephalitis occurs primarily in elderly or immunocompromised patients with a higher incidence than previously estimated. The diagnosis is often delayed; risk factors for unfavorable outcome are age, cerebral vasculitis, and GCS <15.