The nature and effects of trust in social and political institutions have been studied in adults, but few studies have focused on how trust affects the political socialization of children and adolescents, who are in the process of developing their attitudes towards government and other social institutions. Data collected in 1999 from the IEA Civic Education Study of 14-year-olds are used to examine trust at three levels — trust in institutions with which individuals have little or no daily contact (those delegated as representatives in institutions such as the national legislature), trust in institutions with whose representatives individuals interact frequently (schools), and trust in other people. First, levels of these three types of trust are compared in six democracies whose levels of political stability vary (French-speaking Belgium, Bulgaria, Chile, Colombia, England and the United States). Second, correlates of individuals' levels of trust (including school climate and experiences with family) are examined. Third, trust, civic knowledge, school experiences, and family variables are used to predict levels of three types of civic or political engagement (voting, conventional political participation that goes beyond voting, and community participation). Levels of trust relate to the stability of democracy in the countries examined and to participation, suggesting a ‘threshold’ of trustworthiness that a political system needs to establish in order to foster civic and political participation in young people. Additionally, different types of civic engagement are influenced differentially by trust and by other aspects of experience in schools.