BackgroundEarly intervention in psychosis (EIP) services are the dominant service model in the treatment of first-episode psychosis. They are a time-limited intervention and little is known about discharge destinations and outcomes once EIP treatment has concluded.AimsTo understand discharge pathways and predictors of relapse in an EIP service.MethodWe collected data on all patients with an electronic health record treated by EIP services in Oxford Health NHS Foundation Trust in the UK between 12 January 2006 and 7 March 2017 (n = 701). Our primary outcomes were discharge destination at end of treatment and relapse.ResultsMost patients (83.5%) were discharged to primary care. Transfer to secondary care was associated with previous in-patient admissions (odds ratio (OR) = 1.92, 95% CI 1.54–2.39) and longer EIP treatment (OR = 1.04, 95% CI 1.03–1.06). Relapse rate was highest shortly after leaving EIP services. Relapse was associated with transfer to secondary care (hazard ratio (HR) = 2.75, 95% CI 1.75–4.31), higher deprivation (HR = 1.03, 95% CI, 1.01–1.05), a substance misuse disorder (HR = 1.81, 95% CI 1.01–3.26) and a comorbid diagnosis of a personality disorder (HR = 2.96, 95% CI 1.39–6.29).ConclusionsMost patients treated by the EIP service in Oxfordshire did not receive ongoing mental healthcare from secondary mental health services. We identified high deprivation and those with substance misuse problems or personality disorders as EIP populations with a high risk of relapse.Declaration of interestNone.