Objectives Studies assessing emotional distress severity of paediatric oncology patients prior to reaching 5‐year survivorship status have produced inconsistent findings. This cross‐sectional multi‐centre study aimed to determine psychological, demographic, illness and treatment risk factors for emotional distress in this population. Methods Paediatric oncology patients (n = 74), aged 12–18 years, completed the Paediatric Index of Emotional Distress, Self‐Description Questionnaire and Paediatric Quality of Life Inventory. Seventy‐two parents provided background information regarding demographics, diagnoses and treatment protocols in addition to appropriate proxy ratings. Results Hierarchical multiple regression analyses demonstrated that demographic, illness and treatment factors explained little variance in emotional distress. Global self‐concept, global‐generic and cancer‐specific health‐related quality of life (HRQOL) were significant predictors. Provisional multiple regression analyses indicated that pain/hurt, illness‐related worries, communication difficulties and negative self‐views in relation to math abilities, parent relations and opposite‐sex peer relations were risk factors for emotional distress in this sample. Paired‐sample t‐tests and Pearson's moment‐correlation coefficients showed patient and parent reports of patients' self‐concepts and HRQOL were highly consistent. Conclusions This study empirically identified modifiable psychological risk factors for emotional distress prior to 5‐year survivorship status and provided guidance for future interventions. Furthermore, findings suggest that parent reports can provide reliable estimates of patients' self‐concepts and HRQOL. Generalizability of the findings was enhanced by the diversity of the sample studied, in terms of diagnosis and treatment exposure, and the multi‐centre recruitment strategy employed. Nevertheless, the findings should be corroborated by larger, longitudinal studies. Copyright © 2014 John Wiley & Sons, Ltd.