Objective: To determine human tear proteins using nanoliter liquid chromatography coupled to quadrupole orbitrap mass spectrometry (NanoLC-Q-Orbitrap-MS), and perform a bioinformatics analysis of main proteins. Methods: Human tear samples were collected with capillary, transferred to 3 kDa ultrafiltration tubes containing 400 μL of superpure water and centrifuged at 12 000×g for 15 min. Repeated extraction of tear proteins were performed four times, and following digestion with trypsin, the proteins were separated using the Waters NanoAcquity peptide BEH C18 column (1.7 μm, 100 μm×100 mm) and determined using NanoLC-Q-Orbitrap-MS with the mobile phase of 0.1% formic acid aqueous solution and acetonitrile (0.1% formic acid) in the full MS/dd-MS2 mode. The types of proteins were characterized in the Uniprot database using the software Proteome Discoverer version 2.1 and verified using bovine serum albumin. The tear proteins were subjected to gene annotation analysis using the String database. Results: A total of (387±160) human tear proteins were yielded, with a relative standard deviation (RSD) of 4.13%, and there were 25 types of proteins with a relative high abundance, including lipocalin 1, lysozyme and lactoferrin. The peptide sequence coverage of bovine serum albumin was (86.08±2.61)%, with a RSD of 3.03%. The 25 major tear proteins were involved in substance transduction among cells, homeostasis process, negative regulation of the endopeptidase activity, detection of chemical stimulants and humoral immune responses, and the 16 proteins had close interactions. Lacritin, lipocalin 1, lactoferrin, lysozyme and zinc-α 2-glycoprotein, which had a relative high abundance, had close biological connections. Conclusion NanoLC-Q-Orbitrap-MS is stable, reliable and feasible for detection of multiple proteins in tears.