OBJECTIVE: This study was conducted to assess the clinical significance of traumatic brain stem injury (TBSI) reflected on Glasgow Coma Score (GCS) and Glasgow Outcome Score (GOS) by various clinical variables. METHODS: A total of 136 TBSI patients were selected out of 2695 head-injured patients. All initial computerized tomography and/or magnetic resonance imaging studies were retrospectively analyzed according to demographic- and injury variables which result in GCS and GOS. RESULTS: In univariate analysis, mode of injury showed a significant effect on combined injury (p<0.001), as were the cases with skull fracture on radiologic finding (p<0.000). The GCS showed a various correlation with radiologic finding (p<0.000), mode of injury (p<0.002), but less favorably with impact site (p<0.052), age (p<0.054) and skull fracture (p<0.057), in order of statistical significances. However, only GOS showed a definite correlation to radiologic finding (p<0.000). In multivariate analysis, the individual variables to enhance an unfavorable effect on GCS were radiologic finding [odds ratio (OR) 7.327, 95% confidence interval (CI)], mode of injury (OR; 4.499, 95% CI) and age (OR; 3.141, 95% CI). Those which influence an unfavorable effect on GOS were radiologic finding (OR; 25.420, 95% CI) and age (OR; 2.674, 95% CI). CONCLUSION: In evaluation of TBSI on outcome, the variables such as radiological finding, mode of injury, and age were revealed as three important ones to have an unfavorable effect on early stage outcome expressed as GCS. However, mode of injury was shown not to have an unfavorable effect on late stage outcome as GOS. Among all unfavorable variables, radiological finding was confirmed as the only powerful prognostic variable both on GCS and GOS.