The aim of this study was to examine the role of the quality of family and peer relationships in explaining social anxiety in adolescents. The total of 290 students of one high school in Zagreb County participated in the study. The study was conducted durin one class period and participants completed Peer Relationship Scale for Adolescents, Family Relationship Perception Scale, Liebowitz's Social Anxiety Scale for Children and Adolescents and a general socio-demographic data questionnaire. Students mostly percieve their peer relationships as high quality, they percieve their mothers and fathers generaly high on the dimension of emotionality and low on the dimension of control and they also report a low degree of social anxiety. A t-test was conducted indicating significant gender differences in the estimated quality of family relationships and the emotionality of father in the direction of higher results in boys. There were also significant differences in social anxiety, its subscales and the quality of peer relationships, more precisely in the areas of intimacy, trust and insightfulness in the direction of higher results in girls. Results show a significant negative correlation between social anxiety and assessment of the quality of family relationships, the emotionality of the mother and father and the quality of peer relationships and a significant positive correlation with the control of the father. Furthermore, it was also shown that subjective assessment of the quality of family relationships is significantly positively correlated with the quality of family relationships, as is the emotionality of mother and father with the quality of peer relationships. Social anxiety is significantly negatively associated with the subscale of popularity, while associations with subscales of intimacy and trust are significant, but negligible low. A four-step hierarchical regression analysis was conducted, with fear and avoidance as components of social anxiety and overall social anxiety a criteria and with gender, age, quality of family and peer relationships as predictors. The percentages of explained variance range from 31.7% to 35.8%, with more frequent significant predictors of gender, popularity and father emotionality, and somewhat contradictory findings regarding parental control. It can be concluded that family and peer relationships play an important role in explaining adolescent social anxiety and it is necessary to explore the mechanisms in the background in more detail.