The APACHE II (Acute Physiology and Chronic Health Evaluation) system has been widely used as an objective means of predicting outcome in critically ill patients. We prospectively evaluated patients consecutively admitted to the medical intensive care unit to compare the predictive accuracy of APACHE II with clinical assessment by critical care personnel. At the time of admission to the intensive care unit, the house staff and nurse responsible for each patient were asked to estimate the patient's hospital mortality risk. The patient's APACHE II score was calculated and a prediction of the patient's hospital mortality risk was then computed on the basis of this score. A total of 366 patients were studied. Mortality predictions were obtained from 57 physicians and 33 critical care nurses. We were unable to demonstrate a significant difference in the accuracy of APACHE II predictions compared with either physicians' or nurses' predictions. Clinical assessment and APACHE II were both highly predictive of outcome.