Gene expression profiling via RT-qPCR is a robust technique increasingly used in ecotoxicology. Determination and validation of optimal reference genes is a requirement for initiating RT-qPCR experiments. To our best knowledge, this study is the first attempt of identifying a set of reference genes for the freshwater crustacean Gammarus fossarum. Six candidate genes (Actin, TUB, UB, SDH, Clathrin and GAPDH) were tested in order to determine the most stable ones in different stress conditions and to increase the robustness of RT-qPCR data. SDH and Clathrin appeared as the most stable ones. A validation was performed using G. fossarum samples exposed for 15 days to AgNO 3 , silver nanoparticles (AgNPs) 40 nm and gold nanoparticles (AuNPs) 40 nm. Effects on HSP90 were evaluated and data normalized using Clathrin and SDH. A down-regulation of HSP90 was observed when G. fossarum were exposed to AuNPs 40 nm whereas no effects were observed when G. fossarum were exposed to AgNPs 40 nm. This study highlights the importance of the preliminary determination of suitable reference genes for RT-qPCR experiments. Additionally, this study allowed, for the first time, the determination of a set of valuable genes that can be used in other RT-qPCR studies using G. fossarum as model organism. Understanding the mechanisms underlying the effects of stressors on organisms needs sensitive analytical techniques that can cover and link responses observed at different biological levels (from molecular to individual responses). One of the most reliable techniques is the measurement of changes or alterations in gene expression in response to an external stimulus 1. Recent advances in "omics" and bioinformatics methodologies applied in eco-toxicological studies provided a new angle of studying non-model organisms, opening new ways in determining new molecular biomarkers (genes) as an alteration of their regulation may influence the fitness of organisms 2,3. Reverse transcription quantitative polymerase chain reaction (RT-qPCR) is currently described as one of the most reliable techniques to assess these changes due to its effectiveness, sensitivity and reproducibility 1,2,4. This method, which allows studying the expression of a set of selected genes in an organism, requires multiple critical quality controls in order to obtain robust results. This includes RNA purity and integrity control, genomic DNA contamination assessment, evaluation of PCR primer efficiency and specificity and, in case of relative quantifica-tion of gene expression, the identification of suitable reference genes for data normalization 1,5,6. Reference genes are described to be stable regardless of the exposure conditions and overall treatments groups within an experimental design, making them suitable for data normalization of genes of interest. Therefore, their determination is crucial for a good analysis of RT-qPCR results 2,5,7,8. Furthermore, combination of a set of the most stable reference genes is fundamental to avoid misinterpretation of the results that may occur when using a unique reference gene 1. Consequently, a preliminary experimental determination of the optimal reference genes is necessary before each new experimental design 5,8. For this purpose, many software programs and tools have been developed in order to determine the best reference genes. This selection is performed following a ranking method based on their stability across different stress conditions 5,7,9. Unfortunately, many ecotoxicological studies are still using only one, or a set of reference genes selected from previous studies, applying different experimental