Lung cancer is a leading cause of deaths in cancer. Hence, developing early-stage diagnostic tests that are non-invasive, highly sensitive, and specific is crucial. In this study, we investigated to determine whether biomarkers derived from urinary volatile organic compounds (VOCs) can be used to discriminate between lung cancer patients and normal control patients. The VOCs were extracted from the headspace by solid-phase microextraction and were analyzed by gas chromatography time-of-flight mass spectrometry. Nine putative volatile biomarkers were identified as elevated in the lung cancer group. Receiver operating characteristic curve analysis was also performed, and the markers were found to be highly sensitive and specific. Next we used principal component analysis (PCA) modeling to make comparisons compare within the lung cancer group, and found that 2-pentanone may have utility in differentiating between adenocarcinoma and squamous cell carcinomas.