Objective: To screen the prognostic risk long noncoding RNAs (lncRNAs) for pancreatic ductal adenocarcinoma (PDAC) by bioinformatics approaches. Methods: The RNA-Seq Level 2 data and clinical information of PDAC patients were downloaded from The Cancer Genome Atlas (TCGA) database. The sequencing data of the downloaded mRNAs and lncRNAs were re-annotated according to the gene annotation data from NCBI Gene database and GENCODE v7 database. The differentially expressed mRNAs and lncRNAs were screened by using edgeR and limma packages in R. Then, the significantly co-expressed pairs between mRNAs and lncRNAs were obtained by correlation analysis, in which, the mRNAs were considered as target genes of the lncRNAs. After that, the functional modules of the lncRNAs were predicted by functional enrichment analysis of their target mRNAs with the clusterProfiler package in R. Finally, the significant prognostic risk lncRNAs for PDAC were determined by drawing Kaplan-Meier curves of the differentially expressed lncRNAs. Results: After gene re-annotation, the sequencing data of a total of 19 791 mRNAs and 1 623 lncRNAs were obtained, and then, 260 differentially expressed mRNAs and 15 differentially expressed lncRNAs were picked up. From the correlation analysis, 24 significantly co-expressed pairs comprised of 24 mRNAs and 5 lncRNAs were identified. Of them, LINC00857 had 6 target genes that were C1orf116, ESRP1, GPRC5A, LIPH, MAL2 and PLS1, respectively. According to the functional enrichment analysis, the target genes of lncRNA LINC00857 were mainly enriched in phospholipase activity, structural constituent of cytoskeleton, and lipase activity. The results of survival analysis revealed that lncRNA CASC8 and LINC00857 were significantly associated with prognostic risk of PDAC (P=0.0052, P=0.027). Conclusion: CASC8 and LINC00857 are potential prognostic risk lncRNAs for PDAC, and may probably become the novel indictors for prognosis of PDAC in the future.