Background: This study aims to compare mortality prediction capabilities of three different physiological trauma scoring systems (TSS): Revised Trauma Score (RTS); Glasgow Coma Scale, Age, and Systolic Blood Pressure (GAP); and Mechanism, Glasgow Coma Scale, Age, and Arterial Pressure (MGAP).Study design: A descriptive, cross-sectional study of trauma victims admitted to the emergency service between December-2013 and February-2014. Clinical and epidemiological information were gathered at admission and three TSS were calculated: RTS, GAP, and MGAP. The follow-up period to assess length of hospitalization and mortality lasted until August-2014. Two groups were created -survivals (S) and deaths (D) -and compared. P < 0.05 was considered statistically significant. Results:A total of 668 trauma victims were analyzed. The mean age was 37 ± 18 and 69.8% were males. Blunt trauma prevailed (90.6%). The mean scores of RTS, GAP, and MGAP for group S (n = 657; 98.4%) were 7.77 ± 0.33, 22.8 ± 1.7, and 27.4 ± 2.3 respectively (p < 0.001), whereas group D (n = 11, 1.6%) achieved mean scores of 4.57 ± 2.95, 13 ± 7, and 15.5 ± 7 (p < 0.001). Regarding the Receiver Operating Characteristics (ROC) analysis, the areas under the curve were 0.926 (RTS), 0.941 (GAP), and 0.981 (MGAP). The three TSS demonstrated significant mortality prediction capabilities (p < 0.001). There was no statistically significant difference between the three ROC curves (p = 0.138). The MGAP achieved the highest sensitivity (100%), while GAP and RTS sensitivities were 81.8% (59-100%), and 90.9% (73.9-100%) respectively (p < 0.001). The observed specificities were 96.2% (94.77-97.7%) for GAP, 91.6% (89.5-93.7%) for MGAP, and 87.2% (84.7-89.8%) for RTS (p < 0.001). Age (p = 0.049), Glasgow Coma Scale (GCS) (p < 0.001), and trauma mechanism (p < 0.001) were different between the two groups. Conclusion:Most patients were young males and victims of blunt trauma. The three TSS demonstrated reliability regarding mortality prediction. The MGAP achieved the highest sensitivity and GAP was the most specific score, which may indicate a potential use of both as valuable alternatives to RTS. Keywords:Cross-sectional studies, Emergency medical services, Glasgow coma scale, Hospitalization, Mortality, Receiver operating characteristics curve, Sensitivity and specificity, Trauma severity indices, Triage. Resultados: Analisaram-se 668 vítimas de trauma. Registrou-se média de idade de 37±18 anos, 69,8% sendo masculinos. Predominou o trauma contuso (90,6%). Para o grupo S (n = 657; 98,4%), as médias de RTS, GAP e MGAP foram, respectivamente, 7,77 ± 0,33, 22,8 ± 1,7 e 27,4 ± 2,3 (p < 0,001), enquanto o grupo O (n = 11, 1,6%) obteve médias de 4,57 ± 2,95, 13 ± 7 e 15,5 ± 7 (p < 0,001). A análise Receiver Operating Characteristics (ROC) revelou áreas abaixo da curva de 0,926 (RTS),