Recently many genetic mutations that are associated with epilepsy have been identified. The protein NIPA2 (non-imprinted in Prader-Willi/Angelman syndrome region protein 2) is a highly selective magnesium transporter encoded by the gene NIPA2 in which we have found three mutations (p.I178F, p.N244S and p.N334-E335insD) within a population of patients with childhood absence epilepsy (CAE). In this study, immunofluorescence labeling, inductively coupled plasma-optical emission spectroscopy (ICP-OES), MTT metabolic rate detection and computational modeling were utilized to elucidate how these mutations result in CAE. We found in cultured neurons that NIPA2 (wild-type) proteins were localized to the cell periphery, whereas mutant proteins were not effectively trafficked to the cell membrane. Furthermore, we found a decrease in intracellular magnesium concentration in the neurons transfected with mutant NIPA2, but no effect on the survival of neurons. To understand how low intracellular magnesium resulted in hyperexcitability, we built and analyzed a computational model to simulate the effects of mutations. The model suggested that lower intracellular magnesium concentration enhanced synaptic N-methyl-D-aspartate receptor (NMDAR) currents. This study primarily reveals that a selective magnesium transporter NIPA2 may play a role in the pathogenesis of CAE.