BackgroundThere is a growing consensus to reduce the use of restrictive care practices in mental health settings to minimise the physical and psychological complications for patients. However, data regarding restrictive care practice use and factors contributing to variations in the proportion estimates has not previously been synthesised.AimsThis study aimed to synthesise evidence on (1) the pooled proportions of physical restraint, seclusion or chemical restraint in adult mental health inpatients and (2) sources of variability in these proportion estimates.MethodsStudies were identified from Scopus, MEDLINE, PsycINFO, Web of Science, Embase and CINAHL databases following the PRISMA 2020 guidelines. We conducted a meta‐analysis of studies published in English language from 1 January 2010 to 15 August 2022. Binomial data were pooled using a random effect model, with 95% confidence intervals. Meta‐regression was also computed to identify factors that may contribute to variations in the proportion estimates.ResultsA total of 77 studies were included in this meta‐analysis. The pooled prevalence of physical restraint, seclusion and chemical restraint was 14.4%, 15.8% and 25.7%, respectively. Data were heterogeneous across studies (I2 > 99%). Reporting practices and geographical locations contributed to the variability in the reported estimates of restrictive care practices, with studies from Asian countries reporting higher proportions.ConclusionThere appear differences between geographical locations in the proportion of restrictive practices in mental health inpatients; however, this is complicated by how these prevalence data have been measured and defined. Consistency in the reporting of restrictive care practices in mental health is required to make valid comparisons between geographical regions, policy settings and practice innovations.Relevance to Clinical PracticeEfforts are needed to develop training programmes and policy changes to ensure consistency in defining and reporting of restrictive care practices in mental health facilities.Patient/Public ContributionThis is a systematic review that analysed data from previously published studies, and there was no patient/public contribution in this study.Protocol RegistrationThe protocol for this review has been registered to PROSPERO: CRD42022335167.