The aim of this study was to distinguish non-small-cell lung cancer from normal controls and explore the new potential biomarkers of lung cancer by serum proteomics technology. Label-free quantitative liquid chromatography tandem mass spectrometry (1D-LC/MS/MS) analysis was performed on eight non-small-cell lung cancer (NSCLC) serum samples and eight normal controls. The proteomic data we obtained was analyzed by normalized, randomly paired t test and integrated bioinformatic methods, including hierarchical clustering analysis, principal-component analysis, and support vector machine. We obtained 931 proteins with at least two peptides identified from the 16 serum samples, and 62 proteins were differentially expressed between non-small-cell lung cancer patients and normal controls. There were 16 proteins expressed much higher in the lung cancer group than in the controls. Two hundred eight proteins were shared in all 16 serum samples. Through hierarchical clustering analysis and principal-component analysis based on the 62 differentially expressed proteins, we could distinguish non-small-cell lung cancer from the normal controls. The prediction accuracy of non-small-cell lung cancer analyzed by the support vector machine algorithm based on 208 proteins which were shared in all serum samples is 93.75%. Protein expression patterns have changed in the serum of non-small-cell lung cancer patients. Label-free quantitative LC/MS/MS may be a good method to improve the diagnostic accuracy for lung cancer and it can help in discovering the new biomarkers.