To identify the risk factors associated with mortality of trauma victims during hospitalisation in the intensive care unit (ICU).Prospective cohort.Brazilian ICU specialising in the care of trauma victims.The subjects were divided into two groups: survivors and non-survivors. The variables used to compare the groups included demographic and clinical characteristics and illness/injury severity (Acute Physiology and Chronic Health Evaluation [APACHE II], Simplified Acute Physiology Score [SAPS II], Logistic Organ Dysfunction System [LODS], Injury Severity Score [ISS] and New Injury Severity Score [NISS]). The data were analysed using descriptive and inferential statistics and multiple logistic regression analysis.The sample consisted of 200 patients (164 males) with a mean age of 40.7 years. The predominant causes of injury were traffic accidents (57.5%) followed by falls (31.0%). The ICU mortality was 19.0%. Logistic regression analysis revealed that one point on the NISS and SAPS II scores increased the risk of death by 6% and 7%, respectively. In contrast, the risk of dying decreased 4% for each day of ICU hospitalisation.Professionals must use the SAPS II and NISS for the early identification of trauma victims at high risk for death especially during the first days of ICU hospitalisation.