Traffic accidents are one of the main causes of the highest increase in mortality in Indonesia. This problem needs attention to anticipate the fall of the death toll in a traffic accident. So in this study, there are response variables and several predictor variables. The purpose of this study was to find out what factors influence the severity of traffic accident victims in Ambon city based on categories and model the severity of traffic accident victims in Ambon city based on significant factors using the Multinomial Logistic Regression method. In this study, the results obtained are factors that significantly affect the severity of the traffic accident victims are sex variables (X1), age (X2), education (X3) and type of vehicle (X5).