Purpose: Lung cancer (LC) is one of the most common causes of death by neoplasia. The use of metabolomics provides overall information about the metabolic disorders caused by the LC. In this work has been applied a metabolomic approach based on gas chromatography coupled to a mass spectrometry (GC-MS) to determine the metabolomic profile in three biological fluids from patients with LC: serum (S), urine (U) and bronchoalveolar lavage fluid (BALF). Experimental description: Treatment of fluids samples was based on the addition of alcoholic solvents to separate the protein fraction after centrifugation. In the case of urine, urease was used to urea removing. Finally, derivatization by methoximation for the protection of carbonyl groups and silylation of polar groups to convert them to volatile compounds was performed before the injection into GC-MS. Results: A total of 90 serum, 90 urine and 55 BALF samples from patients with LC, patients with lung diseases NO cancer (LD), and healthy controls (C), were analyzed in order to compare the metabolomic profiles of these three types of samples. Multivariate analysis, PLS-DA, showed a clear classification of study groups, indicating the existence of altered metabolites in LC. We identified 22, 18 and 20 perturbed metabolites in LC were identified in S, U and BALF, respectively, involving different metabolic pathways associated with LC. Conclusions: The study of metabolic pathways indicated that the metabolism of glycine, threonine and serine was the most disturbed in LC. In addition, evaluation of specificity and sensitivity of altered metabolites by ROC were applied to characterize potential biomarker of LC.