Objective: To establish a new pancreatic cancer risk assessment model based on the differential expression of m1A-related genes in multiple databases and clinical parameters, and provide theoretical basis for pancreatic cancer treatment and prognosis analysis. Methods: Differential analysis of m1A-related genes was performed in the TCGA-GTEx database. Cox analysis and LASSO regression were used to construct a pancreatic cancer prognosis risk model, and the accuracy and sensitivity of the model were verified by gene expression and clinical data from the GEO database and pancreatic cancer cases in our center. Results: CRLS1 and C7orf50 were selected to form the risk model. The study systematically verified that the model had significant efficacy in indicating the prognosis of pancreatic cancer, and the model combined with tumor mutation burden (TMB) had a stronger prognostic indication ability. Conclusion: A new tool was found to indicate the prognosis of pancreatic cancer patients.