Survival rates amongst children with brain tumours are improving. The increase in successful treatment has been accompanied by a shift in emphasis in psychological research and intervention. Now that cure is an increasingly realistic prospect, the focus of psychology is correspondingly moving towards quality of life in long-term survivors of cancer. A growing body of evidence suggests that cancer survivors are at increased risk of internalising and externalising symptomatology and social problems. There is evidence indicating that survivors of brain tumours are at particular risk, and yet data on psychosocial difficulties in this population are scant. The present study evaluated the extent of psychosocial difficulties in adolescent survivors of brain tumours, and carried out a preliminary investigation of factors associated with outcome on the basis of an a priori conceptual model. In particular, social skills deficits, negative appraisals about the impact of cancer, and social cognitions including expectations of friendship and sense of self as 'different' were hypothesised to be associated with psychosocial difficulties. This study compared the functioning of 21 adolescent cancer survivors, aged 11 to 16 years, to population norms and siblings. All survivors had been treated for brain tumours in the posterior fossa region of the brain at least three years prior to participation in the study, and in this respect can be thought of as a relatively homogeneous group. Psychosocial difficulties, social skills, and cognitions about friendship and illness were evaluated using standardised measures. These included the Strengths and Difficulties Questionnaire, the Matson Evaluation of Social Skills with Youngsters, the Children's Friendship Inventory and the Perceived Illness Experience Scale. In addition, the Thematic Apperception Test (TAT) was used to elicit qualitative information about social relationships for cancer survivors and their siblings. According to both self and parent report, cancer survivors had significantly more psychosocial difficulties than population norms. A significant subgroup had psychosocial difficulties in the clinically significant range. Adolescents surviving cancer did not differ in terms of social skills or expectations of friendship compared to healthy peers. In their responses to the TAT, cancer survivors did not differ from their siblings on perceptions of isolation or social difference. However, they demonstrated more negative perceptions of peer relationships than siblings, and expressed themes of group identity less. Factors associated with poorer psychological outcome included greater levels of social skills deficits and negative appraisals of the ongoing impact of cancer on school and on peer problems. In addition, adolescents who had friendship expectations characteristic of casual, rather than close, friendships were more likely to have higher levels of psychosocial difficulties. Screening for psychosocial difficulties and the perceived impact of cancer should form part of the long-term follow-up of survivors of childhood brain tumours.