Objective To establish a preoperative prediction model which can effectively predict the risk of occult abdominal metastatic disease of pancreatic ductal adenocarcinoma. Methods The clinical data of 986 patients with pancreatic ductal adenocarcinoma who underwent surgery in our department from Sep. 2018 to Dec. 2020 were retrospectively analyzed. The variables with P≤0.2 in univariate analysis were included in binary logistic regression model, and independent predictors of occult abdominal metastatic disease were screened out; finally a nomogram prediction model was established. Then receiver operating characteristic (ROC) curve was used to evaluate the prediction efficiency. Results The incidence of occult abdominal metastatic disease was 8.42% (83/986) in the pancreatic ductal adenocarcinoma patients (593 ［60.14%］ males and 393 ［39.86%］ females). The average age of the patients was (62.40±9.43) years old. The results of univariate analysis showed that occult abdominal metastatic disease was associated with abdominal pain, duration of abdominal pain, total bilirubin, prealbumin, alanine aminotransferase, aspartate aminotransferase, alkaline phosphatase, lactate dehydrogenase, carcinoembryonic antigen, α-fetoprotein, carbohydrate antigen (CA)125, CA19-9, CA724, ascites, tumor size, tumor site, breakthrough of capsule, the invasion of surrounding organs, the degree of invasion of celiac trunk, superior mesenteric artery, splenic artery and splenic vein, suspected metastasis in the No. 9, No. 13, No. 14 and No. 17 lymph nodes, suspected metastasis of retroperitoneal lymph nodes, suspected metastatic tumor and the sites in the liver, abdominal lesions of unknown nature, and history of abdominal surgery (all P＜0.05). Multivariate analysis showed that age, ascites, short cross-sectional diameter of tumor, the invasion of surrounding organs, the degree of invasion of superior mesenteric artery, suspected metastasis in the No. 13 lymph nodes, suspected liver metastases, uncertain lesions in abdominal cavity and history of abdominal surgery were independent predictors of occult abdominal metastatic disease. Based on the above independent predictors, a nomogram prediction model has been constructed, and the area under the ROC curve was 0.783 (P＝0.001). The optimal risk score, sensitivity and specificity of the model were 77.68, 0.650 and 0.787, respectively. Conclusion The nomogram prediction model can help to improve the preoperative diagnosis rate of occult abdominal metastatic disease of pancreatic ductal adenocarcinoma.