Reverse transcription followed by quantitative polymerase chain reaction (qRT-PCR) is a powerful and commonly used tool for gene expression analysis. It requires the right choice of stably expressed reference genes for accurate normalization. In this work, we aimed to select the optimal reference genes for qRT-PCR normalization within different brain areas during the first week following pentylenetetrazole-induced seizures in immature (P20–22) Wistar rats. We have tested the expression stability of a panel of nine housekeeping genes: Actb, Gapdh, B2m, Rpl13a, Sdha, Ppia, Hprt1, Pgk1, and Ywhaz. Based on geometric averaging of ranks obtained by four common algorithms (geNorm, NormFinder, BestKeeper, Comparative Delta-Ct), we found that the stability of tested reference genes varied significantly between different brain regions. The expression of the tested panel of genes was very stable within the medial prefrontal and temporal cortex, and the dorsal hippocampus. However, within the ventral hippocampus, the entorhinal cortex and amygdala expression levels of most of the tested genes were not steady. The data revealed that in the pentylenetetrazole-induced seizure model in juvenile rats, Pgk1, Ppia, and B2m expression are the most stable within the medial prefrontal cortex; Ppia, Rpl13a, and Sdha within the temporal cortex; Pgk1, Ppia, and Rpl13a within the entorhinal cortex; Gapdh, Ppia, and Pgk1 within the dorsal hippocampus; Rpl13a, Sdha, and Ppia within the ventral hippocampus; and Sdha, Pgk1, and Ppia within the amygdala. Our data indicate the need for a differential selection of reference genes across brain regions, including the dorsal and ventral hippocampus.