Background and Aims Intrahepatic cholangiocarcinoma (ICC) is characterized by its insidious onset and high invasiveness, often resulting in patients losing the optimal timing for surgery upon diagnosis, leading to poor 5-year survival rates following surgery. Early prediction of survival benefits from radical resection is crucial. This study was performed to predict survival benefits in ICC patients undergoing radical resection based on preoperative imaging and serum markers to provide guidance and references for clinical decisions regarding the suitability of radical resection.Methods The imaging and serological data of 821 ICC patients who underwent radical resection in 13 tertiary-grade A-class hospitals in China from January 2010 to December 2021 were retrospectively collected. Imaging data included the presence of liver mass, intrahepatic bile duct dilation, portal vein invasion, lymph node invasion, ascites, and stones. Serum markers had hemoglobin, white blood cell count, lymphocyte count, neutrophil count, alpha-fetoprotein (AFP), carcinoembryonic antigen (CEA), carbohydrate antigen 19-9 (CA19-9), CA125, alanine aminotransferase (ALT), total bilirubin (TBIL), albumin (ALB), and prothrombin time (PT). Target variables were selected through univariate and multivariate Cox regression analysis, and using these variables, a CoxPH model was constructed, and a nomogram was also visualized. Survival curves were plotted using Kaplan-Meier analysis to validate the relationship between the scores and patient outcomes. The model's predictive performance was assessed through receiver operating characteristics (ROC) and calibration curves.Results The independent prognostic factors were imaging indicators of ascites, intrahepatic bile duct dilation, lymph node invasion, and serum markers with CEA >5 μg/L, CA19-9 >37 U/mL, and CA125 >40 U/mL (all P<0.05). The CoxPH model built using the 6 variables demonstrated that patients in the high-risk group identified by the model had significantly lower 1-, 3-, and 5-year survival rates compared to the low-risk group (all P<0.05). The model exhibited good discrimination and effectiveness through the constructed nomogram. ROC curves showed that the area under the curve (AUC) for the model's predictions at 1, 3, and 5 years were 0.711, 0.721, and 0.782, respectively, surpassing the predictive efficacy of individual markers.Conclusion The prognostic model composed of preoperative CA125, ascites, intrahepatic bile duct dilation, lymph node invasion, CEA, and CA19-9 effectively stratifies patients into high- and low-risk groups and provides accurate individualized predictions of survival benefits following radical resection in ICC patients. This model offers guidance for clinical decisions regarding the suitability of radical resection.