We compared placebo and dexamethasone (initial dose, 3 mg/kg; total, 11.4 mg/kg per 48 h) in a double-blind trial involving 10 stuporous and 28 comatose patients with cerebral malaria. Patients were 18 mo to 42 y of age (geometric mean, 10.2 y), and the 19 patients in each group were comparable on admission. All patients received intravenous quinine therapy. Four patients (21%) in each group died. There were no significant differences between the placebo- and dexamethasone-treated groups in time until patients became afebrile (median, 51 vs. 19 h), the level of consciousness became normal (mean, 80 vs. 83 h), or parasitemia was cleared (mean, 2.1 vs. 3.4 d) or in the incidence of complications. Coma or hyperparasitemia (greater than or equal to 5% of erythrocytes parasitized) at the time of admission and hypoglycemia at any time during hospitalization were significantly correlated with a fatal outcome, which was not improved by using dexamethasone. We conclude that high-dose dexamethasone is not indicated for treating cerebral malaria.