Using gas chromatography-mass spectrometry (GC-MS), a new metabolic profiling method was established to assess the levels of non-esterified fatty acids (NEFAs) and esterified fatty acids (EFAs) in plasma. The extraction method was simple and robust without removing protein process. With this method 25 fatty acids (FAs), both EFAs and NEFAs, can be recognized simultaneously with only 10 mu L plasma. 15 of the 25 can be precisely quantified. The method was validated and then applied into clinical metabonomics research. Five clinical groups including 150 cases were involved. The relationship between FA levels and diabetic mellitus (DM) as well as diabetic nephropathy (DN) pathology was speculated. Furthermore, the possible pathological causes and effects were discussed in detail. Potential biomarkers (p value < 0.01) were screened with Student's t-test. With the application of partial least squares-discriminant analysis (PLS-DA), different stages were distinguished. The result may be useful for the pathology study of metabolic syndromes, and may also be helpful for monitoring the progression of DM and DN. (C) 2011 Elsevier B.V. All rights reserved.