Traumatic brain injury (TBI) is a leading cause of death and disability. A prognostic model is useful to calculate the probability of TBI outcome. It is a statistical model that includes two or more prognostic factors as independent variables and TBI outcome as the dependent variable. In present study the outcome is dichotomous, recovery or death. Objectives: - To develop a prognostic model for the prediction of in hospital mortality in patients with traumatic brain injury (TBI) admitted to the Neurosurgery Intensive care unit (ICU). Materials & Methods: - The demographic data of patients admitted after a diagnosis TBI in ICU were reviewed. A total of 411 patients with TBI were admitted to ICU during the study period. The endpoint of noting the outcome was ICU discharge. Results: - The in-hospital mortality was 15.3%. On multivariate analysis the patients age, conscious level after injury, pupillary reactivity, Alcohol consumption, ENT bleed and hypertension found statistically associated with mortality. These significant variables were further considered for Logistic Regression Analysis. The overall correct prediction was 89.5% (368/411). To predict proportion of deceased & recovered equalised and high, ROC curve analysis was carried out. It revealed the new cut off probability of 0.097 with sensitivity 0.833 and specificity 0.789. The area under receiving operating characteristics (ROC) curve was 0.852. Conclusion: - On Logistic Regression analysis patients age, conscious level after injury, pupillary reactivity, Alcohol consumption, ENT bleed and hypertension were highly associated with ICU outcome. As model is excellent for predicting in hospital mortality in patients with TBI, the probability tables for various combinations of independent variables will help clinicians and relatives in knowing the probable outcome.