Contrasting the well-described effects of early intervention (EI) services for youth-onset psychosis, the potential benefits of the intervention for adult-onset psychosis are uncertain. This paper aims to examine the effectiveness of EI on functioning and symptomatic improvement in adult-onset psychosis, and the optimal duration of the intervention. 360 psychosis patients aged 26-55 years were randomized to receive either standard care (SC, = 120), or case management for two (2-year EI, = 120) or 4 years (4-year EI, = 120) in a 4-year rater-masked, parallel-group, superiority, randomized controlled trial of treatment effectiveness (Clinicaltrials.gov: NCT00919620). Primary (i.e. social and occupational functioning) and secondary outcomes (i.e. positive and negative symptoms, and quality of life) were assessed at baseline, 6-month, and yearly for 4 years. Compared with SC, patients with 4-year EI had better Role Functioning Scale (RFS) immediate interaction estimate = 0.008, 95% confidence interval (CI) = 0.001-0.014, = 0.02 and extended social network (interaction estimate = 0.011, 95% CI = 0.004-0.018, = 0.003) scores. Specifically, these improvements were observed in the first 2 years. Compared with the 2-year EI group, the 4-year EI group had better RFS total ( = 0.01), immediate ( = 0.01), and extended social network ( = 0.05) scores at the fourth year. Meanwhile, the 4-year ( = 0.02) and 2-year EI ( = 0.004) group had less severe symptoms than the SC group at the first year. Specialized EI treatment for psychosis patients aged 26-55 should be provided for at least the initial 2 years of illness. Further treatment up to 4 years confers little benefits in this age range over the course of the study.