Background. The PEWS (Pediatric Early Warning Score) is always used for prognosis of critical incidents in children at hospital. However, recording of various vital deviations may be advantageous. In this regard, it is relevant to study their prognostic value for risk of mortality estimation. Objective. The aim of the study was to estimate the frequency of vitals deviations from reference ranges in patients with low and high risk of mortality. Methods. We have studied medical records of patients aged from 1 month to 17 years old hospitalized in intensive care unit (ICU). Mortality prediction was counted with PIM3 score. Low and high risk of mortality groups were created due to the results of cluster analysis. The low risk group was made up of patients with mortality prediction rate < 37%, high risk group > 57%. Heart rate, respiratory rate, oxygen saturation in arterial blood, body temperature, systolic and diastolic blood pressure and deviations of all these values from reference ranges on presentation to ICU and further daily (every 2 hours) during the first 5 days (or less in case if the patient was transferred from the ICU earlier) were analysed. Results. Medical records of 66 patients (including 10 (15%) with a high risk of mortality) have been studied. The low and high risk groups were comparable by sex, age, spectrum of diseases, length of patient stay in the ICU. The lethal outcome in ICU was registered in two cases: both patients with high risk of mortality (> 50%) due to PIM3 score. The frequency of vitals deviations from reference ranges on presentation to ICU and in the following 5 days (except for HR on the first day) in risk groups did not differ statistically. Conclusion. Registration of only vitals deviations from reference ranges in the child during first 5 days in ICU does not allow us to predict high risk of mortality measured by PIM3 score.