Abstract Background The high risk of psychosis among migrants is often attributed to social stressors in the host country. We examined whether the relative risk of psychosis among migrants is low on arrival and increases thereafter. Methods In this cohort study, first-generation immigrants to the Netherlands, aged 10 years and older ( N = 1 281 678), were matched by birth year and sex to 2 542 313 native-born Dutch controls. The first occurrence of psychosis after arrival was established using data on dispensing of antipsychotic medication (APM) (during 2006–2017) and on insurance claims for treatment of psychosis (2011–2016). The Incidence Rate Ratios (IRRs) for migrants compared to controls were estimated by year since arrival. Results The IRR of APM was 0.22 (95% CI 0.21–0.24) in the year of arrival (‘year 1’) and increased gradually to 1.39 (1.19–1.62) after 10 or more years. The IRR of an insurance claim increased from 0.57 (0.51–0.62) to 1.87 (1.38–2.55) in year 5. Among migrants from sub-Saharan Africa, the IRR of an insurance claim was already high in year 1 [2.46 (1.95–3.11)], especially when aged 10–20 years at arrival [6.09 (2.93–12.64)]. Among migrants from other non-Western countries, the IRR was already significantly increased in year 2 [1.28 (1.03–1.59)]. Conclusions The relative risk of psychosis among migrants was generally low at arrival and increased thereafter. The increased IRRs in the early years after arrival among those from non-Western countries indicate that for these groups certain risk factors are already relevant shortly after arrival.