Objective: To assess systematically the effect on quality of survival (QoS) of the addition of chemotherapy to craniospinal irradiation (CSI) for treatment of primitive neuro-ectodermal tumour (PNET). Methods: We undertook a cross-sectional multi-informant questionnaire to assess QoS in UK children previously enrolled in the International Society for Paediatric Oncology PNET3 randomized controlled trial (RCT) of CSI (35Gy plus 2OGy boost to posterior fossa) versus CSI plus neo-adjuvant chemotherapy (Vincristine, Carboplatin, Etoposide, and Cyclophosphamide). When comparing the survivors assessed in the present follow-up study, pre- and post-surgical evaluation of neurological function, undertaken at the time of diagnosis, had shown no difference between children in the two treatment arms. Age-appropriate measures of outcome, as assessed by patients, parents, and health professionals, were compared between the treatment arms, using x2, student t-test, or Mann-Whitney U test, in 127/170 (75 %) of survivors at a mean (SD) age of 15.4 (4.0) years and 7 (2.25) years from diagnosis. These are given here for 69 12-17-year-olds for whom the following assessments were employed: (1) neurological assessment by a medical examination form and also by responses of parents to a questionnaire; (2) self-report and proxy- (i.e. principal carer) rated scores for both health status (Health Utilities Index MU [HUI3]), and behavioural problems (the Strengths and Difficulties Questionnaire [ SDQ]); and (3) quality of life (Paediatric Quality of Life questionnaire [PedsQL]). Results: In 12-17-year-olds allocated to the CSI + chemo arm, relative to the CSI alone arm: (1) Neurological function was more impaired, particularly with respect to physical restriction of activity (n=69, x2 = 10.78, p=0.001); (2) HUI scores were significantly lower both on self-assessment (n=63, U=300.5, p=0.005) and proxy assessment (n=63, U=344, p=0.04), explained by differences in the domain of cognition and dexterity; (3) 'Total difficulties' scores on the SDQ were higher (n=63, t=2.l, p=0.04) on parental assessment with more emotional symptoms. There was a similar trend on self-assessment; and (4) PedsQL scores were significantly lower both on self-assessment (n=63, U=345.0, p=0.045) and proxy-assessment (n=63, t=-2.829, p=0.006) with lower scores for physical health. Differences between treatment arms, showed a similar pattern of significant differences in the 18-24-year-olds (n=38) but not in the 6-11-year-olds (n=20). There were clear relationships between HUI, SDQ, and PedsQL group scores. Conclusions: The addition of chemotherapy to 'standard dose' CSI may adversely affect health status, behaviour, and quality of life. The effect on these measures of reducing the dose of CSI and simultaneously adding chemotherapy (e.g. in the current PNET4 RCT) requires further investigation