BACKGROUND: Few data are available on intensive care unit (ICU) patient populations and critical care medicine practices in developing countries. METHODS: This prospective study evaluated differences in patient characteristics, ICU practice, and outcome between the ICUs of a Mongolian 400-bed tertiary university hospital (MonICU) and an Austrian 429-bed secondary hospital (AutICU). Demographics, chronic health status, clinical parameters, disease and therapeutic severity scores, and outcome were documented for all patients admitted to the two ICUs during a period of four and a half months. Standard tests and multiple regression analysis were used for statistical analysis. RESULTS: A total of 203 critically ill patients were admitted to MonICU and 257 to AutICU. MonICU patients had fewer chronic diseases than AutICU patients (0.9 +/- 0.8 vs. 2.7 +/- 1.5, P < 0.001) but more frequently suffered from tuberculosis (2.5% vs. 0%, P = 0.01) and more frequently had never been medically examined before ICU admission (10.8% vs. 0%, P < 0.001). Admission diagnoses differed both in type and relative proportions in the two ICUs (P < 0.001). Admission of MonICU patients was more frequently unplanned (69% vs. 50.2%, P < 0.001), and although disease was more severe in these patients they received fewer therapeutic interventions than the AutICU patients. Overall mortality was higher in the MonICU patients (19.7 vs. 6.2%, P < 0.001). CONCLUSIONS: Patient characteristics and ICU practices varied significantly between the two ICUs. Mortality was substantially greater at MonICU, particularly among patients suffering from multiple-organ dysfunction. Strategies to improve the care of critically ill patients at MonICU should address both system- and staff-related problems, improve acceptance of the ICU service among physicians of other disciplines and upgrade the training of ICU staff.