In order to validate the proposed method we focused on three metabolites of interest with various functional groups and polarities including CH3-malonic acid (MMA: biomarker of methylmalonic acidemia), 3-hydroxy-3-methyl-glutaric acid (3-OHMGA: biomarker of 3-hydroxy-3-methylglutaric acidemia), and phenylpiruvic acid (PhPA: marker of phenylketonuria). While these three metabolites can be considered as representative of organic acids classically determined by 1D-GC, they cannot be representative of new detected metabolites. Thus, we also focused on quinolic acid (QUIN), taken as an example of biomarker not detected at basal levels with the classical 1D GC-qMS method. In order to obtain sufficient recoveries for all tested compounds, we developed a sample preparation protocol including a step of urea removal followed by two extraction steps using two solvents of different polarity and selectivity. Recoveries with the proposed method reached more than 80% for all targeted compounds and the linearity was satisfactory up to 50 μmol/L. The CVs of the within-run and within-laboratory precisions were less than 8% for all tested compounds. The limits of quantification (LOQs) were 0.6 μmol/L for MMA, 0.4 μmol/L for 3-OHMGA, 0.7 μmol/L for PhPA, and 1 μmol/L for QUIN. The LOQs of these metabolites obtained by a classical GC-MS method under the same chromatographic conditions were 5 μmol/L for MMA, 4 μmol/L for 3-OHMGA, 6 μmol/L for PhPA while QUIN was below the limit of detection. As compared to 1D-GC, these results highlight the enhanced detectability of urine metabolites by the 2D-GC technique. Our results also show that for each new detected compound it is necessary to develop and validate an appropriate sample preparation procedure.