The research problem has been conceptualized as an examination of the possibility to predict socially unacceptable behaviour by using resilience as a predictor. In this context, self-reported socially unacceptable behaviour has been examined, while the resilience has been analysed with regard to two aspects in which it can be conceptualized. The first one is Ego-resiliency which is defined as a personality trait, and it is related to adaption during the inhibition or expression of emotions, in an interdependent relation to the environment in which the person is. The second aspect is resilience defined through intrapersonal and interpersonal capacities of person which affect adaptation and tolerance to stress and negative life experiences. The research has been conducted by using convenience sample of 384 respondents of high school final-year students, from Republika Srpska (BiH). The hierarchical regression analysis has been used, while the applied measurement instruments were the following: RSA - Resilience Scale for Adults, Ego-resiliency scale, the scale of anti-social behaviour, as well as scales that measure dimensions of Aggressiveness and Conscientiousness from the Big five +2 instrument. The results of the hierarchical regression analysis showed resilience as a statistically significant predictor of socially unacceptable behaviour (R = .40, R2 = .160, F(4,318) = 15.09, p = .000). When it comes to individual predictors, Conscientiousness (β = -.24, p = .000) and Aggressiveness (β = .19, p = .000) are still dominant (after the two measures of resilience are added to the model in the second step), which is followed by the resilience (RSA; β = -.16, p = .012). It is important to note that Ego-resiliency has not been identified as a statistically significant predictor of socially unacceptable behavior (β = .06, p = .342). The results contribute to the understanding of antisocial behavior causes and consequently contribute to its prevention.