Ninety-seven acutely ill patients with bronchial asthma were enrolled in a double-blind, placebo-controlled, randomized trial of intravenous methylprednisolone (125 mg), given on presentation in the emergency room in addition to standard emergency treatments for asthma. Subjective and spirometric indexes of the severity of the asthma were similar on entry into the study in all patients, but only 9 of 48 patients (19 percent) treated with methylprednisolone required hospital admission, as compared with 23 of 49 patients (47 percent) in the control group (P less than 0.003). Our results suggest that prompt use of glucocorticoids in the emergency treatment of severe asthma can prevent significant morbidity, reduce the number of hospitalizations, and effect substantial savings in health care costs.