This paper tests the individual-level and household-level human, social, and cultural capitalpredictors of the frequency of civic engagement (volunteer, political and organisationalinvolvement) among 4,760 youths aged 10 to 15 years in the UK. Human capital resources donot consistently predict youth civic engagement. Initial social class differences in civicengagement disappear after controlling for cultural capital resources. Self-esteem is positivelycorrelated with organisational involvement but negatively so with political activity. In terms ofsocial capital resources, number of friends is positively associated with volunteering and politicalactivity. Cultural capital resources are the most relevant as aesthetic tastes and parental rolemodelling both positively predict of all three forms of civic engagement. Policy should focus onengaging young lower social class urban White males and encouraging parental civicinvolvement or mentoring programs that supplement engagement.