Background: Adolescent attachment relationships with parents and peers and the sense of connectedness with the schools attended have been established as salient predictors of psychological well-being. Few studies, however, have assessed the relative importance of each attachment or connectedness relationship and how they interrelate to influence mental health outcomes. Method: A total of 203 adolescents (11-16 years) completed self-report measures of parental and peer attachment (Inventory of Parental & Peer Attachment-Revised; Gullone & Robinson, 2005); school connectedness (Psychological Sense of School Membership; Goodenow, 1993); conduct problems, emotional symptoms and prosocial behaviour (Strengths and Difficulties Questionnaire; Goodman, 1997). Results: Multiple regression analyses demonstrated that more insecure parental attachment (although not peer attachment or school connectedness) predicted conduct problems and emotional difficulties. Peer attachment and school connectedness were significant predictors of prosocial behaviour, whereas parental attachment was not. A mediational analysis revealed that peer attachment and school connectedness both mediate the relationship between parental attachment and prosocial behaviour. No significant moderation effects of either peer attachment or school connectedness on the relationship between parental attachment and mental health outcomes were found. Conclusions: Different attachment and connectedness relationships, although related, predict adolescent mental health outcomes in distinct ways. Improving parental attachment may have particular salience in reducing negative behaviours such as conduct problems and emotional difficulties, whereas improving peer attachment and school connectedness could be important for the display of prosocial behaviour.