Differential protein expression profiles in the serum samples from patients with lung adenocarcinoma may be associated with glycosylation during cancer development. In this study, we used various glycoproteomic approaches to investigate the different glycoproteomic profiles of human normal and lung adenocarcinoma serum samples and to investigate putative altered glycoprotein biomarkers. In our preliminary screening, FITC-labeled lectin staining was used for the detection of specific glycoprotein profiles. wheat germ agglutinin (WGA) lectin had the highest level of specific binding to glycoproteins in both samples. We enriched for glycoproteins in the serum samples using WGA lectin affinity and then performed co-immunoprecipitation with anti-haptoglobin and 2-DE, 2-D difference in-gel electrophoresis and MS analyses. From these analyses, we identified 39 differentially expressed proteins, including 27 up-regulated proteins and 12 down-regulated proteins. Bioinformatics tools were used to search for protein ontology, category classifications and prediction of glycosylation sites. In addition, three up-regulated glycoproteins (adiponectin, cerulolasmin and glycosylphosphatidyl-inositol-80) and two down-regulated glycoproteins (cyclin H and Fyn) that were found to be correlated with lung cancer development were validated by Western blot analysis. We suggest that these altered glycoproteins may be useful as biomarkers for lung cancer development and progression.