Objective: To investigate the predictive value of silent information regulator 2 (SIRT2) and microRNA-28-5p (miR-28-5p) expression levels in cholangiocarcinoma tissues on the prognosis of radical resection. Methods: A total of 280 patients with cholangiocarcinoma who underwent radical resection were prospectively selected and divided into training set (n=196) and validation set (n=84) according to the ratio of 7∶3. The general data and prognosis of training set and validation set were compared. All patients were followed up for 2 years after operation, and the prognosis was statistically analyzed. According to the prognosis, the training set was divided into poor prognosis group (n=81) and good prognosis group (n=115). The general data, expression levels of SIRT2 mRNA and miR-28-5p in cancer tissues were compared between the two groups. Logistic regression was used to analyze the influencing factors of poor prognosis in patients with cholangiocarcinoma. Receiver operating characteristic curve (ROC) was used to analyze the value of SIRT2 mRNA and miR-28-5p in predicting poor prognosis in patients with cholangiocarcinoma. Based on SIRT2 mRNA and miR-28-5p, a nomogram model was constructed to analyze the predictive value of the nomogram model, and internal validation was performed in the validation set. Results: In the training set, the expression level of SIRT2 mRNA in cancer tissues was higher than that in normal tissues adjacent to the cancer, while the expression level of miR-28-5p in cancer tissues was lower than that in normal tissues adjacent to the cancer (P<0.05). The proportion of Child-Pugh B classification, tumor maximum diameter >4 cm, lymph node metastasis, moderate to poor tissue differentiation, TNM stage III-IV, age-adjusted Charlson comorbidity index (ACCI) >4, carbohydrate antigen 19-9 (CA19-9) >40 kU/L, and carcinoembryonic antigen (CEA) >5 μg/L, as well as the expression level of SIRT2 mRNA in cancer tissue of patients in poor prognosis group, were higher than those in the good prognosis group (P <0.05). The proportion of patients undergoing R0 resection and postoperative chemotherapy, as well as the expression level of miR-28-5p in cancer tissue, were lower than those in the group with good prognosis (P<0.05). Lymph node metastasis, moderate to poor tissue differentiation, TNM stage III-IV, ACCI>4, and SIRT2 mRNA expression in cancer tissue were independent risk factors for poor prognosis in patients with cholangiocarcinoma. R0 resection, postoperative chemotherapy, and miR-28-5p expression in cancer tissue were independent protective factors (P<0.05). The area under the curve (AUC) for predicting poor prognosis in patients with cholangiocarcinoma using SIRT2 mRNA and miR-28-5p in cancer tissue was 0.819 and 0.807, respectively (P<0.05). The AUCs for predicting poor prognosis in patients with cholangiocarcinoma using the nomogram model in the training set and validation set were 0.942 and 0.921, respectively, indicating high predictive performance and high consistency and fit between predicted and actual outcomes. Conclusion: The expression levels of SIRT2 mRNA and miR-28-5p in cancer tissues of patients with cholangiocarcinoma can be used as predictors of poor prognosis after radical resection. The nomogram model based on SIRT2 mRNA and miR-28-5p in cancer tissues can provide reliable clinical basis for clinical identification of patients with high risk of poor prognosis.