To evaluate whether pretreatment albumin-globulin ratio (AGR) can be used as a biomarker for predicting the prognosis of patients with urothelial carcinoma (UC).We systematically searched PubMed, Web of Science, China National Knowledge Infrastructure (CNKI), Google Scholar and Cochrane Library; the search time was up to May 2022. Stata 16.0 was used for data processing and statistical analysis.We identified 12 studies with 5,727 patients from 317 unique citations during the meta-analysis. Our results suggested that a low AGR before treatment was significantly associated with poor overall survival (OS) [hazard ratio (HR) = 1.99, 95% confidence interval (CI) = 1.45-2.75, P < 0.001], cancer-specific survival (CSS) [HR=2.01, 95% CI = 1.50-2.69, P < 0.001] and recurrence-free survival (RFS) [HR=1.39, 95% CI = 1.12-1.72, P = 0.002]. Furthermore, we defined different subgroups according to ethnicity, cancer type, cut-off value, sample size and stage. Similar prognostic outcomes for OS and CSS were observed in most subgroups. However, for subgroup of stage, the low pretreatment AGR only predicted the poor survival of patients with non-metastatic UC.Our meta-analysis revealed that the AGR before treatment could be used as a predictive biomarker to indicate the prognosis of UC patients during clinical practice, especially in patients with non-metastatic UC.Copyright © 2022 Xia, Fu, Li, Wu, Niu, Xu, Wang, Yuan and Tang.