OBJECTIVE: To value of glucocorticoid (GC) in treatment for patients with acute respiratory distress syndrome (ARDS) was evaluated. METHODS: The clinical data from all patients with ARDS in medical ICU (MICU) during May 2000 to Aug. 2002 were collected. They were divided into two groups, GC and non-GC groups, in order to compare their age, sex, acute physiology and chronic health evaluation II (APACHE II) score, PaO2/FiO2, Qs/Qt, level of positive end-expiratory pressure(PEEP), mortality and dead reason of death, depending on whether GC was given or not. In cases with administration of GC, the dosage, as well as duration of treatment was analyzed in terms of the overcome. RESULTS: There were 77 cases totally, among them 60 cases were of GC group and 17 of non -GC. Their age, sex, APACHE II score, PaO2/FiO2, Qs/Qt, use of artificial ventilation and its duration, level of PEEP, and the extent of relief from hypoxemia showed no significant differences between two groups (P>0.05). Even the mortality for patients who were treated with GC was higher than those without (71.7% vs. 52.9%), though there was no statistically significant difference (P>0.05). The percentage of patients died primarily of ARDS was low in both groups (7.0% and 11.1%). The age, APACHE II score and underlying diseases for non-survivors were older and higher than survivors (P<0.001 or P<0.005) and their duration of staying in ICU was shorter than the latter (P<0.05). The mortality of patients who were given GC before or during 24 hours of the establishment of the diagnosis of ARDS (66.7% and 68.2%) was lower than those who were given GC 24 hours after the diagnosis of ARDS (90.0%). CONCLUSION: GC could be one of effective treatments for ARDS, and it should be given without hesitation when refractory hypoxemia and shock were found.