Pancreatic adenocarcinoma (PAAD) is a malignant tumor with one of the highest associated mortality rates worldwide, and a 5-year survival rate of <5%. Fibroblast growth factors (FGFs) serve important roles in numerous cellular functions, and dysregulation of FGFs contributes to various cancer types. However, there are few reports on the function of FGFs in PAAD. The Assistant for Clinical Bioinformatics database, Gene Expression Profiling Interactive Analysis, Kaplan-Meier plotter and Tumor Immune Estimation Resource were utilized to perform the protein-protein interaction network, functional enrichment, univariate Cox regression, least absolute shrinkage and selection operator (LASSO) Cox, differential expression, prognostic value and immune cell infiltration analyses of FGFs in patients with PAAD. Immunohistochemistry (IHC) was used to verify the predictive value of the model. A total of 22 FGF genes were identified. Based on the results of LASSO Cox regression analysis, a total of six genes, including FGF2, FGF8, FGF9, FGF13, FGF17 and FGF22, were selected for the establishment of the prognostic gene signature. High transcriptional levels of FGF17 and FGF22 were significantly associated with long overall survival. The expression of FGFs was associated with the infiltration of various immune cells. According to univariate and multivariate analyses, FGF2 and FGF8 may be useful independent prognostic biomarkers for the prognosis of patients with PAAD. IHC demonstrated that FGF2 and FGF8 were more highly expressed in PAAD tissues compared with that in normal tissues. The present findings offer a novel understanding for the selection of FGF prognostic biomarkers in PAAD.