A growing number of studies have regarded the preoperative serum albumin-to-globulin ratio (AGR) as a prognostic indicator of urothelial carcinoma (UC) following radical surgery. However, a pooled analysis of AGR's effect on UC prognosis was still insufficient. Up to January 2022, a systematic search was conducted using PubMed, Embase, Web of Science, and Cochrane Library. Stata SE software was applied in this study. The reviewers collected the hazard ratio (HR) with 95% confidence interval (CI) for overall survival (OS), cancer-specific survival (CSS), recurrence-free survival (RFS), progression-free survival (PFS), and metastasis-free survival (MFS). A total of 9,002 patients from 12 retrospective studies were included in this analysis. The results showed that preoperative serum AGR was significantly associated with the OS (HR = 1.85, 95%CI = 1.43 to 2.39), CSS (HR = 2.38, 95%CI = 1.69 to 3.34), RFS (HR = 1.64, 95%CI = 1.29 to 2.08), PFS (HR = 2.16, 95%CI = 1.43 to 3.27), and MFS (HR = 3.00, 95%CI = 1.63 to 5.53) of patients with UC following radical surgery. Sensitivity analysis indicated the stability of the results. Subgroup analysis revealed that preoperative low AGR was seen as a risk factor for OS (HR = 1.90, 95%CI = 1.34 to 2.69), CSS (HR = 2.13, 95%CI = 1.40 to 3.26), and RFS (HR = 1.60, 95%CI = 1.24 to 2.07) in upper tract urothelial carcinoma (UTUC), but it was only a risk factor for CSS (HR = 2.95, 95%CI = 1.14 to 7.60) in bladder cancer (BC). Besides, preoperative AGR cut - value ≤ 1.4 could not be deemed as a stable prognostic indicator for RFS (HR = 2.07, 95%CI = 0.71 to 6.04) in UC. However, the predictive ability of AGR cut - value > 1.4 was stable. All in all, preoperative low AGR was considered as a risk factor for UC. AGR level can be regarded as a prognostic indicator for OS, CSS, and RFS in UTUC but only for CSS in BC. AGR greater than 1.4 can be a great cut-off value for predicting the prognosis of UC patients with radical operation.