Benefits of capillary electrophoresis to provide a comprehensive snapshot of multiple metabolites in biological samples have been exploited. Afterwards, multivariate statistical methods can be employed in order to mine additional information from the data. Urine fingerprints of control and diabetic rats have shown the clear effects of an antioxidant treatment on diabetic animals, which were not seen in controls, in a rapid, simple and cost-effective way without identifying a single marker. The procedure involves the measurement of samples with a relatively inexpensive tool such as CE-UV, without any previous treatment other than filtration and the application of chemometric tools [PCA (principal components analysis) and PLS-DA (partial least squares discriminant analysis)]. Data pre-treatment of electrophoretic profiles (alignment, normalization and baseline correction) has shown to be key for further chemometric treatment. Once developed, the methodology can easily be applied for a rapid in vivo screening of extracts with potential in vitro activity. Classification was supported by that produced after PCA and PLS-DA of target variables obtained with selectively designed, time and reagent consuming methods.