We enrolled 200 infants and older children with bacterial meningitis in two prospective double-blind, placebo-controlled trials to evaluate the efficacy of dexamethasone therapy in addition to either cefuroxime (Study 1) or ceftriaxone (Study 2). Altogether, 98 patients received placebo and 102 received dexamethasone (0.15 mg per kilogram of body weight every six hours for four days). At the beginning of therapy, the clinical and demographic characteristics of the patients in the treatment groups were comparable. The mean increase in the cerebrospinal fluid concentration of glucose and the decreases in lactate and protein levels after 24 hours of therapy were significantly greater in those who received dexamethasone than in those who received placebo (glucose, 2.0 vs. 0.4 mmol per liter [36.0 vs. 6.9 mg per deciliter], P less than 0.001; lactate, 4.0 vs. 2.1 mmol per liter [38.3 vs. 19.8 mg per deciliter], P less than 0.001; and protein, 0.64 vs. 0.25 g per liter [64.0 vs. 25.3 mg per deciliter], P less than 0.05). One patient in the placebo group in Study 1 died. As compared with those who received placebo, the patients who received dexamethasone became afebrile earlier (1.6 vs. 5.0 days; P less than 0.001) and were less likely to acquire moderate or more severe bilateral sensorineural hearing loss (15.5 vs. 3.3 percent; P less than 0.01). Twelve patients in the two placebo groups (14 percent) had severe or profound bilateral hearing loss requiring the use of a hearing aid, as compared with 1 (1 percent) in the two dexamethasone groups (P less than 0.001). We conclude that dexamethasone is beneficial in the treatment of infants and children with bacterial meningitis, particularly in preventing deafness.