Objective: To enable further understanding of how cognitive deficits and psychopathology impact psychosocial functioning in first-episode psychosis patients, we investigated how these clinical domains were associated with psychosocial problems at baseline, and if they predicted psychosocial functioning at 12-months follow-up. Also, we tested whether baseline cognitive deficits were stronger predictors of change in psychosocial functioning in the first year after diagnosis than baseline psychotic symptoms. Methods: 162 first-episode psychosis (FEP) patients completed assessment of eight neurocognitive- (attention, problem solving, speed of processing, verbal fluency, verbal learning, visual learning, working memory, general cognition) and four social cognitive subdomains (emotion perception, theory of mind, social knowledge, social cognitive biases) and psychopathology (positive- and negative symptoms, depression and anxiety) at baseline. Psychosocial functioning (work/study, relationships, self-care, disturbing behaviour and general psychosocial functioning) was assessed at baseline and 12-months follow-up. Backward regression models were computed to test our hypotheses. Results: At baseline, psychosocial functioning was associated strongest with positive- and negative symptoms of all assessed clinical subdomains, followed by neurocognition and social cognition. In contrast, psychosocial functioning at 12-months follow-up was not predicted by psychotic symptoms, but rather by neurocognition, social cognition and depression. Change in social functioning in the first year after baseline was predicted by positive and negative symptoms, but to a similar degree by neurocognition and social cognition. Conclusions: Whereas psychotic symptoms show marked impact on functioning at illness onset, cognitive deficits appear to be more accurate longitudinal predictors of psychosocial problems and functional recovery in the early stages of psychosis.