The relationship between cognition and outcome in people with schizophrenia has been established in studies that, for the most part, examined chronic patients and were cross-sectional in design. The purpose of this study was to analyze the relationships between neurocognitive variables assessed at illness onset and functional outcome in a longitudinal design. An additional area of interest was whether the severity of negative symptoms would predict outcome independently from neurocognitive variables or whether there would be an overlap in their predictive power. The authors administered a comprehensive cognitive battery and clinical assessments to 99 subjects who were in their first episode of illness and analyzed the relationship of cognition and symptom severity at intake with community outcome after an average follow-up period of 7 years. Verbal memory, processing speed and attention, and the severity of negative symptoms at intake were related to subsequent outcome. Global psychosocial functioning was predicted by negative symptoms and attention. Verbal memory was the significant predictor of the degree of impairment in recreational activities. Impairment in relationships was predicted by negative symptoms and memory, whereas attention and negative symptoms were predictive of work performance. There was an overlap in the variance in outcome explained by cognitive variables and negative symptoms. Verbal memory and processing speed and attention are potential targets for psychosocial interventions to improve outcome. Results from cross-sectional or chronic patient studies do not necessarily correspond to the findings of this prospective first-episode study in which cognition appears to explain less of the variance in outcome.