BACKGROUND: Restrictive interventions are used in the treatment of some people with severe mental disorders such as psychosis - including psychiatric intensive care unit (PICU) admission, seclusion and restraint. Early Intervention in Psychosis (EIP) service input may improve outcomes in psychosis, but it is unclear whether specific components of EIP care reduce the need for restrictive practice. AIMS: To examine associations between EIP care components, demographic characteristics and restrictive interventions. METHOD: We conducted a retrospective cohort study of 14 874 people who used EIP services in England, using linked data from the National Clinical Audit of Psychosis and the Mental Health Services Data Set. We examined associations between EIP components and time to PICU admission (primary outcome) alongside seclusion/physical restraint/injected chemical restraint/requests for police assistance (secondary outcomes), using multilevel Cox regression, adjusting for demographic factors and clustering by service. RESULTS: Higher hazards of restrictive interventions were observed among men, younger people and several minority ethnic groups. Individuals eligible for clozapine who were not offered it (hazard ratio 1.51, 95% CI 1.20-1.91) or refused it (hazard ratio 1.46, 95% CI 1.02-2.10) had higher hazards of PICU admission than those not eligible, whereas those who were eligible for clozapine and received it did not. There was weaker evidence of similar effects on hazards of physical restraint and seclusion. Receipt of CBT for psychosis was associated with reduced hazards of PICU admission (hazard ratio 0.80, 95% CI 0.67-0.95) and physical restraint (hazard ratio 0.68, 95% CI 0.47-0.98). Substance use was associated with increased hazards of PICU admission and requests for police assistance, although substance use interventions appeared to partially mitigate this. CONCLUSIONS: Marked demographic disparities exist in the use of restrictive practice. Specific EIP care components may be associated with reductions. Strengthening evidence-based EIP provision and addressing structural inequalities may support progress towards less coercive and more equitable care.