Glycoproteins constitute an important part of the human proteome as they regulate e.g. cell-cell communication and recognition. These proteins are of great biological importance and commonly found in the cell membrane. However, due to their expected low concentration level, an extensive and selective isolation is necessary prior to miniaturized LC-MS/MS analysis.In this study immobilized hydrazide on beads has been explored to selective isolation and enrichment of N-linked glycoproteins. The isolation was carried out on both protein and peptide level. In short, the glycans in the glycoproteins were oxidized to aldehydes with sodium periodate and reacted with hydrazide beads overnight. The unattached proteins were subsequently removed by extensive washing and centrifugation prior to reduction, alkylation and PNGase F deglycosylation of the glycoproteins. The deglycosylated proteins were digested using trypsin and analysed by monolithic reversed phase liquid chromatography Quadrupole-Orbitrap mass spectrometry (RP LC-Q-Orbitrap-MS/MS). The method was used on standard proteins as well as BxPC-3 pancreatic cancer primary cell line samples. With isolation on peptide level, the proteins were digested prior to connection with immobilized hydrazide. The proteins were identified using the SEQUEST™ search engine.Secondary interaction and non-selective interactions assumed to be between protein-bead and protein-protein were observed, and may prevent the glycoproteins from being enriched. Although there were less secondary interactions for the peptide level isolation, the presence of only one peptide left the protein identification uncertain. Only one of the three glycoproteins in the standard samples was identified with the glycomotif indicating possible non-efficient deglycosylation. For the BxPC-3 samples isolated on protein level only 4 % of the identified proteins were found to be glycosylated and no glycomotif was identified in the peptide level isolation.The BxPC-3 cell line samples were additionally enzymatically digested to investigate the number of proteins one can expect to identify in various numbers of cells and to determine the repeatability of sample preparation and injection repeatability. The proteins identified overlapped with 35 % and 65 % for the sample preparation and injection repeatability, respectively.