Schizophrenia is a debilitating mental disorder. Currently, the lack of disease biomarkers to support objective laboratory tests constitutes a bottleneck in the clinical diagnosis of schizophrenia. Here, a gas chromatography-mass spectrometry (GC-MS) based metabolomic approach was applied to characterize the metabolic profile of schizophrenia subjects (n = 69) and healthy controls (n = 85) in peripheral blood mononuclear cells (PBMCs) to identify and validate biomarkers for schizophrenia. Multivariate statistical analysis was used to visualize group discrimination and to identify differentially expressed metabolites in schizophrenia subjects relative to healthy controls. The multivariate statistical analysis demonstrated that the schizophrenia group was significantly distinguishable from the control group. In total, 18 metabolites responsible for the discrimination between the two groups were identified. These differential metabolites were mainly involved in energy metabolism, oxidative stress and neurotransmitter metabolism. A simplified panel of PBMC metabolites consisting of pyroglutamic acid, sorbitol and tocopherol-α was identified as an effective diagnostic tool, yielding an area under the receiver operating characteristic curve (AUC) of 0.82 in the training samples (45 schizophrenia subjects and 50 healthy controls) and 0.71 in the test samples (24 schizophrenic patients and 35 healthy controls). Taken together, these findings help to develop diagnostic tools for schizophrenia.