In this study, we describe the optimization of a Hydrophilic Interaction Liquid Chromatography coupled to mass spectrometry (HILIC-MS) method for the evaluation of 14 metabolites related to the de novo synthesis of pyrimidines (dnSP) while using multivariate analysis, which is the metabolic pathway for pyrimidine nucleotide production. A multivariate design was used to set the conditions of the column temperature, flow of the mobile phase, additive concentration, gradient rate, and pH of the mobile phase in order to attain higher peak resolution and ionization efficiency in shorter analysis times. The optimization process was carried out while using factorial fractional designs, Box–Behnken design and central composite design while using two zwitterionic columns, ZIC-p-HILIC and ZIC-HILIC, polymeric, and silica-based columns, respectively. The factors were evaluated while using resolution (R), retention factor (k), efficiency of the column (N), and peak height (h) as the response variables. The best optimized conditions were found with the ZIC-p-HILIC column: elution gradient rate 2 min, pH 7.0, temperature 45 °C, mobile phase flow of 0.35 mL min−1, and additive (ammonium acetate) concentration of 6 mM. The total analysis time was 28 min. The ZIC-p-HILIC LC-MS method yielded satisfactory results for linearity of calibration curves, limit of detection (LOD), and limit of quantification (LOQ). The method has been shown to be appropriate for the analysis of dnSP on samples of tomato plants that were infected with Phytophthora infestans.