Introduction. According to the World Health Organization, suicide is the fourth leading cause of death between the ages of 15 and 29. Students are the predominant part of this age group. At the same time, medical students have a higher prevalence of suicidal thoughts and suicide attempts compared to their peers, which makes it necessary to focus attention on this problem. The aim of this study is to investigate the predictors of suicidal risk among medical university students. Materials and Methods. The study was conducted on the basis of the Northern State Medical University (Arkhangelsk). Students of the first and fifth years of medical and pediatric faculties were included. A sociological method was used to assess suicidal risk and to identify the level of suicidal intentions formation. For data collection we used A. G. Shmelev's Suicide Risk Questionnaire. Analysis of variance using descriptive statistics was performed, the significance level (p) was set at 0.05 for all tests. The significance of the influence of external factors on the degree of suicidal risk was assessed by conducting an age-adjusted analysis of covariance (ANCOVA). The obtained integral risk score served as a basis for constructing a logistic regression model to predict the level of suicidal tendency in students. Results. 441 people were included in the study. Distribution of respondents according to the level of suicidal risk: none – 0,9%, low – 22,7%, decreased – 44,0%, average – 25,1%, increased – 6,6%, high – 0,7%. Statistically significant influence on the increase of suicidal risk was: belonging to the faculty of pediatrics (OR=2,11); fifth year (OR=2,68); having a single-parent family (OR=3,45); living in a dormitory (OR=2,9); tense family environment (OR=12,91); tense group environment (OR=9,1); having a relative with a psychiatric illness in the 1st or 2nd line of descent (RPI) (OR=7,29). The maximum chance of having high suicidal risk was for students with unfavorable family and group environment – the odds ratio was 23,86 and 233,8 for individuals with no and having a RPI, respectively. The trend of the model showed a 2,36-fold increase in the odds ratio. The area under the ROC curve for the model was 0,78 and the pseudo-R-square was 25%. Conclusions. Individual models have proven that the three factors that have the greatest impact on increasing the level of suicidal risk are: a tense family situation, a tense group situation and the presence of RPI. The pseudo-R-square equal to 25% makes it necessary to search for other possible suicide risk factors that improve the model metrics. The detection of additional factors, which will increase the area under the ROC curve, can serve as the basis for the development of a screening method for early detection of high suicide risk.