A comprehensive metabonomic method, in combination with fingerprint analysis and target analysis, was performed to reveal potential mechanisms of berberine action in the treatment of patients with type 2 diabetes and dyslipidemia. Serum samples of 60 patients before and after treatment with either berberine or placebo were collected. Ultra-performance liquid chromatography and quadrupole time-of-flight tandem mass spectrometry (UPLC Q-TOF MS) coupled with pattern recognition analysis were used to identify changes in global serum metabolites. Compared with placebo, patients before and after berberine treatment could be separated into distinct clusters as displayed by the orthogonal signal correction filtered partial least-squares discriminant analysis (OSC-PLS-DA) score plot, which indicated changes in circulating metabolites after berberine treatment. Among them, free fatty acids changed markedly. These were further quantified by UPLC combined with single quadrupole mass spectrometry (UPLC SQ MS). There was a highly significant decrease in the concentrations of 13 fatty acids following berberine administration. 10 fatty acids also differed statistically from placebo. These results suggest that berberine might play a pivotal role in the treatment of type 2 diabetes through down-regulating the high level of free fatty acids and that comprehensive metabonomic measurements are potentially very useful for studying the mechanisms of action of traditional Chinese medicines.