OBJECTIVE:Increased employment duration has been associated with change in performance on specific neurocognitive domains in populations with schizophrenia, but not in first-episode psychosis. The aim of this exploratory study was to examine whether employment duration over 18 months is associated with neurocognitive outcomes over 18 months among individuals with first-episode psychosis. METHOD:Eighty-eight young people with first-episode psychosis completed a neurocognitive battery at baseline and 18 months. Setwise (hierarchical) multivariate linear regressions were used to examine predictors of change in neurocognitive performance over 18 months. Total hours employed over 18 months were entered after accounting for age, gender, premorbid IQ, and negative symptom change scores. RESULTS:Total hours employed was significantly associated with change in Symbol Digit Modalities Test raw score (p = .020), Letter-Number Sequencing scaled score (p = .016), Digit Span total scaled score (p = .047) and Rey Complex Figure Test delayed recall raw score (p = .016) over 18 months, after controlling demographic characteristics, premorbid IQ, and changes in negative psychotic symptoms. CONCLUSIONS AND IMPLICATIONS FOR PRACTICE:Total hours worked over 18 months was associated with small improvements on one test of processing speed and one test of working memory. However, total hours worked over 18 months was also associated with decline on one test of attention and working memory and visual organization and memory. The findings implicate that work alone may not be entirely effective in changing neurocognitive functioning for young people with first-episode psychosis. (PsycINFO Database Record (c) 2018 APA, all rights reserved).