© 2015 Cambridge University Press. Individuals identified as at ultra-high risk (UHR) for psychosis are at risk of poor functional outcome regardless of development of psychotic disorder. Studies examining longitudinal predictors of poor functioning have tended to be small and report only medium-term follow-up data. We sought to examine clinical predictors of functional outcome in a long-term longitudinal study. Method. Participants were 268 (152 females, 116 males) individuals identified as UHR 2-14 years previously. A range of clinical and sociodemographic variables were assessed at baseline. Functioning at follow-up was assessed using the Social and Occupational Functioning Assessment Scale (SOFAS). Results. Baseline negative symptoms, impaired emotional functioning, disorders of thought content, low functioning, past substance use disorder and history of childhood maltreatment predicted poor functioning at follow-up in univariate analyses. Only childhood maltreatment remained significant in the multivariate analysis (p <0.001). Transition to psychosis was also significantly associated with poor functioning at long-term follow-up [mean SOFAS score 59.12 (s.d. = 18.54) in the transitioned group compared to 70.89 (s.d. = 14.00) in the non-transitioned group, p <0.001]. Childhood maltreatment was a significant predictor of poor functioning in both the transitioned and non-transitioned groups. Conclusions. Childhood maltreatment and transition to psychotic disorder independently predicted poor long-term functioning. This suggests that it is important to assess history of childhood maltreatment in clinical management of UHR individuals. The finding that transition to psychosis predicts poor long-term functioning strengthens the evidence that the UHR criteria detect a subgroup at risk for schizophrenia.