Objective: Bioinformatics methods were used to analyze key genes in cervical lymph node metastasis of papillary thyroid carcinoma (PTC). Methods: Microarray data of PTC GSE60542 was searched from Gene Expression Omnibus (GEO) database of National Center for Biotechnology Information (NCBI). The GEO2R online analysis tool was used to screen differentially expressed genes between PTC and paracarcinoma tissues (differentially expressed gene set A) and differentially expressed genes between PTC and cervical metastatic lymph nodes (differentially expressed gene set B). The differentially expressed gene set C of the intersection of differentially expressed gene set A and differentially expressed gene set B was further screened. And bioinformatics tools including David, String and cBioPortal were applied to analyze the biological function of differentially expressed gene set C and their encoded proteins interaction. Then, we performed survival analysis on key genes. Results: By analyzing the GSE60542 microarray data, a total of 1 086 differentially expressed gene A, 194 differentially expressed gene B, and 39 differentially expressed gene C were obtained. GO analysis showed that these genes are mainly involved in the molecular function of oxidoreductase activity, protein homodimerization activity, growth factor binding and involved in the biological processes of cell proliferation, thyroid hormone generation, negative regulation of neuron apoptotic process, gamma-aminobutyric acid signaling pathway. Through the online cBioPortal analysis website, the gene CCL21, which had a great influence on the survival time of patients, was initially screened. Conclusion: Microarray combined with bioinformatics methods can effectively analyze the differentially expressed genes between PTC tissues, paracarcinoma tissues, and cervical metastatic lymph node tissues, which may provide guidance for molecular mechanisms research of PTC pathogenesis.