Objective: To determine the prognostic factors of the functional outcome of patients surgically treated for acute epidural hematomas. Methods: Two hundred patients who consecutively underwent neurosurgery for acute epidural hematomas over the past 9-year period were studied. Clinical characteristics, radiologic findings, and the time intervals with regard to treatment course were investigated to determine the interactions between all these factors and functional outcome. Results: Functional outcome showed a significant correlation with preoperative consciousness state, Glasgow Coma Scale score, pupillary sizes, and motor posturing (chi(2) test, p < 0.05). Functional outcome correlated with the period of brain hernia tion, the length of time of the operation, as well as the period of hospitalization (chi(2) test, p < 0.05), but not with the length of time of craniotomy decompression relative to the length of time from the injury until admission. The radiologic findings of the associated brain injury, the size and the density of the clot, the degree of the brain shift, and the obliteration of the basal cisterns significantly correlated with functional outcome (chi(2) test, p < 0.05), whereas no significance was attributable to skull fracture. Multivariate analysis indicated that the following four factors independently correlated with functional outcome: (1) associated brain injury, (2) best motor response, (3) hematoma volume, and (4) period of hospitalization (chi(2) test, p < 0.05). A combination of the four factors led to the prediction of the functional outcome with 91% accuracy (1.5% falsely pessimistic predictions and 7.5% falsely optimistic prediction) and 82.1% at over 90% confidence level, These four parameters correlated significantly with preoperative neurologic deterioration (chi(2) test, p < 0.05). Conclusion: This study identifies the risk factors involved in the functional outcome of patients who underwent surgical treatment for acute epidural hematomas, Our results indicate that associated brain injury plus best motor response are the optimal set of two prognostic indicants, with 87% correct predictions and 70.1% at over a 90% confidence level, Prevention of in-hospital neurologic deterioration would improve the patients' functional outcome with a resultant unfavorable recovery rate ranging from 11.5% to 17%.