摘要The treatments for metastatic urothelial carcinoma (mUC) have advanced substantially since 2016. Prognostic tools have been used to inform clinical trial designs and treatment decisions. Historically, prognostic tools were developed for mUC based on older clinical trials involving cytotoxic chemotherapy. As novel therapies emerged, there are studies investigating prognostic factors in the era of immune checkpoint inhibitors (ICI), antibody-drug conjugates, and targeted therapies. This review aims to highlight prognostic factors in mUC and their potential in clinical decision-making and research. In the setting of chemotherapy, patient performance status, site of metastatic burden, and specific laboratory findings were found to have prognostic value in mUC. In the era of ICI, newer models identified variables such as neutrophil to lymphocyte ratio, platelet count, and lactate dehydrogenase to also have potential prognostic value. In addition to clinical biomarkers, molecular biomarkers, such as PD-L1 assay and fibroblast growth factor receptor 2 and 3 genomic testings, may have promising prognostic and predictive implications. Current methods of identifying clinical and molecular prognostic factors involve clinician insight. As large complex datasets emerge, machine learning and artificial intelligence may help data analysis and detect important prognostic features. With careful validation, such machine learning-based strategies may help create more robust prognostic and/or predictive models in the future.