The prognostic significance of preoperative anemia (PA) has been identified in various malignancies. However, its predictive role in urothelial carcinoma (UC) remains controversial. The aim of this study was to investigate the prognostic value of PA in UC patients. We performed a meta-analysis of the association between PA and survival outcome in UC patients. Electronic databases were searched up to June 30, 2016. Study characteristics and prognostic data were extracted from each included study. Cancer-specific survival (CSS), recurrence-free survival (RFS), and overall survival (OS) were pooled using hazard ratio (HR) with corresponding 95% confidence intervals (CI). Herein, 12 studies comprising 3815 patients were included in the meta-analysis. There were 1593 (41.76%) patients in the PA group and 2222 (58.24%) in the control group. The overall pooled HRs of PA for CSS, RFS, and OS were significant at 2.21, (95% CI: 1.83–2.65, Pheterogeneity = 0.49, I2 = 0%), 1.87 (95% CI: 1.59–2.20, Pheterogeneity = 0.22, I2 = 28%), and 2.04(95% CI: 1.76–2.37, Pheterogeneity = 0.36, I2 = 9%) respectively. Stratified analyses indicated that PA was a predictor of poor prognosis based on ethnicity, sample size, tumor T stage, G grade, lymphovascular invasion (LVI), concomitant carcinoma in situ (CIS), and follow-up values. Our findings show that PA has negative prognostic effects on the survival outcome (CSS, RFS, and OS) in UC patients and can serve as a useful and cost-effective marker to aid prognosis prediction.