Admission hyperglycemia and glucose variability were associated with mortality in critically ill patients, but data on trauma patients are to date scarce and heterogeneous. We assessed the prognostic role of ICU death of admission and peak glycemia and glucose variability (indicated by the standard deviation of mean glucose levels and the coefficient of variation of glucose) in 252 patients consecutively admitted for trauma in our ICU (January 1, 2016-December 31, 2018). The in-ICU mortality rate was 17% (43/252). When compared to patients who died during ICU stay, survivors were younger (p = 0.001), more frequently males (p = 0.002), with a lower incidence of hypertension (p = 0.023). Higher values of SAPS II, SOFA and ISS were observed in nonsurvivors (p < 0.001, p < 0.001, p < 0.001, respectively). Survivors exhibited significantly lower values of admission glycemia (p = 0.001), peak glycemia (p = 0.002) and mean glucose values measured during the first 24h since ICU admission (p = 0.001). Glucose variability was significantly higher in nonsurvivors, as indicated by higher values of SD and CV (p = 0.001 and p = 0.001, respectively). At multivariate regression analysis, admission glycemia (Model 1), peak glycemia (Model 2) and glucose variability (Model 3 and 4) were independent predictors for in-ICU mortality. Our findings indicate that not only admission glycemia but also peak glycemia and glucose variability show a correlation with in-ICU mortality in trauma patients.