To learn more about the treatment of herpes simplex encephalitis with vidarabine, we conducted an uncontrolled study of 132 patients referred to 22 hospitals because of suspected disease. All had a brain biopsy and were started on vidarabine, but only 75 were diagnosed by isolation of virus from a brain-biopsy specimen. Cumulative mortality in the latter group was 39 per cent at one year. Other than therapy, levels of consciousness and age were the major variables that influenced outcome. Of 23 patients under 30 years of age who were lethargic at the initiation of therapy, two died and 16 returned to normal. Of 26 patients over 30 years of age who were lethargic at the outset, nine died and 10 returned to normal. Semicoma and coma were associated with worse outcomes, especially in older patients. Our data suggest that outcome is improved with treatment; they provide more support for the use of brain biopsy to diagnose the infection and indicate a need for better therapy. (N Engl J Med. 1981; 304:313–8.) IN a double-blind, placebo-controlled study, 1 vidarabine (adenine arabinoside) reduced mortality due to herpes simplex encephalitis and gave no evidence of drug toxicity. The placebo-controlled investigation was terminated because ethical-review groups thought the data indicated a reduced death rate with treatment. On the basis of these findings, vidarabine was licensed for treatment of this disease. The statistical significance of reduced mortality in that study was questioned, 2 and morbidity could not be determined because of the small number of surviving patients with confirmed disease. Later, the study was criticized on ethical grounds because of the inclusion of placebo controls and the use. © 1981, Massachusetts Medical Society. All rights reserved.