Introduction: Trauma is an important cause of mortality [1, 2]. Researchers are looking for optimal death/survival predictive models in trauma population. One way is to validate traumatic scores for different medical systems [1]. The aim of our study was to validate the New Injury Severy Score (NISS) in severe trauma (ST) from Moldavian population. Material and methods: The retrospective study enroled 467 severe injuried patients (NISS[15), admitted in Reanimatology department of Institute of Emergency Medicine. The regression analysis estimated NISS predictive power for ST adjusted for age and gender. Results: Omnibus Test of Model Coefficients (v2 (df = 2) = 154.310 p\0.001). The Nagelkerke R Square was 51.1%. Hosmer-Lemeshow test indicated an nonsignificant result (v2 (df = 8) = 4.478, p = 0.784). Variables in equation: NISS B = 0.142, p\0.001, OR = 1,14 (95%CI 1.11–1.17) and Age B = 0.058, p\0.001, OR = 1.06 (CI95% 1.04–1.08). Gender was nonsignificant. Sensibility = 81.3% (52 of 64 cases), specificity = 81.9% (327 of 403 cases), mean validation = 81.2% (cut-off 0.123). Area under the ROC curve = 0.900, 95% CI (0.857, 0.944), p p = 1/(1 ? e^(9.179-0.58*Age-0.142*NISS)), where p – death probability in ST; e (exponent)-constant equals to 2.71828 Conclusions: The NISS score, according to model caracteristics, has been validated for the ST patient’s from the Moladavian medical system. Proposed model can/has to be improved in prospective research. Comparative evaluation of different traumatic scores or national models elaboration follows. References: 1. de Munter, L., Polinder, S., Lansink, K.W.W., et al.: Mortality prediction models in the general trauma population: a systematic review. Injury 48(2), 221–229 (2017) 2. Kunitake, R.C., Kornblith, L.Z., Cohen, M.J., et al.: Trauma Early Mortality Prediction Tool (TEMPT) for assessing 28-day mortality. Trauma Surg. Acute Care Open 3(1), e000131 (2018)