Reverse transcription of extracted cellular RNA combined with real-time PCR is now an established method for sensitive detection and quantification of specific mRNA level changes in experimental models of neurological diseases. To neutralize the impact of experimental error and make quantification more precise, normalization of test gene data using data from a constantly expressed gene, a reference gene that is tested along with the test gene, is required. There is no single gene constantly expressed under all experimental conditions. For a given set of conditions or a given disease model, identification of an unaffected reference gene is necessary. In this report, we present our findings from evaluation and validation of the genes encoding hypoxanthine guanine phosphoribosyl transferase 1 (HPRT1) and glyceraldehyde phosphate dehydrogenase (GAPDH) as individual reference genes in mRNA level studies involving four murine neurological disease models. We find both genes are suitable as a reference gene with these four models, provided quantification of subtle changes are avoided. We furthermore demonstrate that above a certain threshold of test mRNA level changes and given high quality RNA processing, normalization to total RNA alone provides for equally reliable quantitative mRNA level results.