BACKGROUND: Cognitive deficits have been well documented in individuals with bipolar disorder (BD) after the first episode of mania. However, little is known about the presence of such deficits prior to the initial manic episode. METHODS: Participants were recruited from a cohort of 416 young people who were at ultra-high risk (UHR) for psychosis and were followed up between 4 and 13 years later. The current report is of 16 participants who developed BD over a mean follow-up period of 8.2 years (UHR-BD). Baseline demographic, clinical and neurocognitive assessment scores were compared with those of 46 age and gender matched UHR subjects who did not transition to psychosis or BD over the follow-up period (UHR-NT) and 66 healthy comparison subjects. RESULTS: UHR-BD subjects had lower global functioning at baseline compared with UHR-NT subjects. There were no significant differences between UHR-BD and UHR-NT subjects on baseline demographic and neurocognitive characteristics. UHR-BD subjects had lower test performance than HC on picture completion, Trail-Making Tests and measures of global intelligence. LIMITATIONS: Small sample size, limited and variable neurocognitive tests utilised and the confounding effects of psychotic symptoms might have impacted on the ability to detect meaningful clinical and neurocognitive differences. CONCLUSIONS: In this exploratory study, neurocognition in young people who later develop BD is similar to those of subjects who are at a high risk for psychotic disorders, but there may be certain neurocognitive markers that distinguish this group from unaffected and healthy young people.