Background: In psychotic patients, functional outcome is associated with several clinical, neuropsychological and developmental factors. In a prospective study design, we aimed to predict functional outcome in patients with first-episode psychosis (FEP) and clinical high risk to psychosis (CHR) by sociodemographic, clinical, neuropsychological and premorbid factors, as well as by follow-up symptomatology.Methods: Altogether, 130 FEP and 60 CHR patients were recruited and their Global Assessment of Functioning (GAF) was assessed at baseline (T0) and when nine months (T1) and 18 months (T2) was elapsed from baseline examination. Sum of follow-up GAFT1 and GAFT2 (F-GAF) was predicted by background characteristics and factorised dimensions of the Structured Interview for Prodromal Syndromes (SIPS) symptoms, Premorbid Adjustment Scale (PAS), Trauma and Distress Scale items and neuropsychological tests, and by psychotic, depression and anxiety symptoms at T1 and T2.Results: F-GAF was predicted by PAS Sociability and School Performance, SIPS Disorganised Thinking, work and marital status in FEP patients. In CHR patients, only SIPS Emotional Poverty predicted significantly F-GAF. In them, follow-up Depression/Anxiety symptomatology is associated strongly with poor F-GAF. The effects of PAS Sociability and School Performance on F-GAF were mediated via SIPS Emotional Poverty in both FEP and CHR patients.Conclusions: In patients with FEP, poor premorbid adjustment, Disorganised Thinking, poor baseline work status and single marital status predict poor follow-up functioning. In both FEP and CHR patients, the effect of premorbid adjustment on follow-up functioning is mediated via Emotional Poverty. In patients with CHR, basic disturbances are affective disorders.