To increase metabolome coverage in global LC-MS metabolomics, often both reversed-phase liquid chromatography (RPLC) and hydrophilic-interaction liquid chromatography (HILIC) are implemented in parallel. However, there is a lack of consensus in the literature on the best HILIC stationary phase to employ for global metabolomics of human biological fluids. The objective of this study was to compare in detail the performance of two commonly employed HILIC phases: zwitterionic sulfobetaine ZIC-HILIC stationary phase and an underivatized silica HILIC stationary phase. During method development, the effect of salt concentration in the mobile phase was also investigated, and 5 mM ammonium acetate was selected. The stationary phases were evaluated using a mixture of 37 polar standards covering a range of logP values (-10 to 3.73), molecular weights (59-776 Da), charges (15 anions, 11 cations, and 11 neutral) as well as 17 lipid standards to understand phospholipid behaviour on the two stationary phases. The criteria used for the comparison included the quality of the chromatographic peak shape, adequate analyte retention, peak separation capability, and metabolite coverage. The zwitterionic ZIC-HILIC column provided better chromatographic performance over the silica stationary phase with 14 standards achieving good quality peaks compared to the 7 with the silica column. Only 2 standards were undetected with the ZIC-HILIC column compared to the 14 undetected with the silica column. In human plasma, 1966 and 1650 metabolites were observed on the ZIC-HILIC column in positive and negative electrospray ionization (ESI) respectively. On the silica HILIC column, 1773 and 2028 metabolites were observed in positive and negative ESI respectively, showing comparable performance of the two phases. Next, the effect of adding 10 mM ammonium phosphate to the samples to improve the analyte peak shape and metabolite coverage was investigated for both ZIC-HILIC and silica HILIC. In contrast with recently reported results for pZIC-HILIC, there was no clear evidence that ammonium phosphate addition was beneficial for human plasma samples. In conclusion, ZIC-HILIC provided better chromatographic performance for polar plasma metabolomics than underivatized silica in terms of chromatographic peak shape and chromatographic resolution, while maintaining comparable metabolite coverage. The addition of ammonium phosphate to human plasma was not beneficial for either of the two stationary phases.