This study performed untargeted metabolomics for plasma samples from 40 coronary heart disease patients and 43 healthy controls by high-performance liquid chromatography coupled with mass spectrometry technology to find a set of effective biomarkers for CHD diagnosis and prognosis. The discriminating metabolites were extracted and analyzed by univariate and multivariate analysis methods. We found five metabolites (1-naphthol, 2-naphthol, methylitaconate, N-acetyl-d-glucosamine 6-phosphate and l-carnitine) contributing to the separation of CHD patients from healthy controls, and a subset of two metabolites in these five were identified as potential plasma biomarkers for CHD diagnosis. Major metabolic pathways associated with these potential biomarkers included nicotinate and nicotinamide metabolism, protein glycosylation, lipid metabolism and fatty acid metabolism. In addition, two potential biomarkers (GlcNAc-6-P and l-carnitine) were found be to be associated with intestinal microflora, indicating that intestinal microflora may be related to the metabolism and progression of CHD.