Objective: Evidence regarding global trends in length of stay for patients with anorexia nervosa is limited. The aim of this study is to conduct a meta-analysis examining the length of stay for anorexia nervosa patients globally, and to investigate moderators of the variance.Method: Medline, EMBASE and PsycINFO were searched for studies published up to January 2019. Two independent reviewers assessed the eligibility of each report based on predefined inclusion criteria. A meta-analysis was performed to calculate a pooled mean length of stay using the random-effects model. Subgroup analyses and meta-regression were conducted to explore potential sources of heterogeneity between studies.Results: Of 305 abstracts reviewed, 71 studies met the inclusion criteria, generating 111 datasets. The pooled mean length of stay (95% confidence interval) was 76.3 days (73.3, 79.4) using the random-effects model, with marked variation across countries and time. Heterogeneity between studies was explained by age and admission body mass index. Conclusions:This meta-analysis found that although clinical features do contribute to length of admissions for anorexia nervosa, there are also global and temporal variations. Future research should provide an in-depth analysis of why and how this variation exists and what the impact is on the well-being of people with anorexia nervosa.