Encephalitis caused by Japanese encephalitis virus occurs in annual epidemics throughout Asia, making it the principal cause of epidemic viral encephalitis in the world. No currently available vaccine has demonstrated efficacy in preventing this disease in a controlled trial. We performed a placebo-controlled, blinded, randomized trial in a northern Thai province, with two doses of monovalent (Nakayama strain) or bivalent (Nakayama plus Beijing strains) inactivated, purified Japanese encephalitis vaccine made from whole virus derived from mouse brain. We examined the effect of these vaccines on the incidence and severity of Japanese encephalitis and dengue hemorrhagic fever, a disease caused by a closely related flavivirus. Between November 1984 and March 1985, 65,224 children received two doses of monovalent Japanese encephalitis vaccine (n = 21,628), bivalent Japanese encephalitis vaccine (n = 22,080), or tetanus toxoid placebo (n = 21,516), with only minor side effects. The cumulative attack rate for encephalitis due to Japanese encephalitis virus was 51 per 100,000 in the placebo group and 5 per 100,000 in each vaccine group. The efficacy in both vaccine groups combined was 91 percent (95 percent confidence interval, 70 to 97 percent). Attack rates for dengue hemorrhagic fever declined, but not significantly. The severity of cases of dengue was also reduced. We conclude that two doses of inactivated Japanese encephalitis vaccine, either monovalent or bivalent, protect against encephalitis due to Japanese encephalitis virus and may have a limited beneficial effect on the severity of dengue hemorrhagic fever.