The use of reverse transcription quantitative PCR technology to assess gene expression levels requires an accurate normalization of data in order to avoid misinterpretation of experimental results and erroneous analyses. Despite being the focus of several transcriptomics projects, oaks, and particularly cork oak (Quercus suber), have not been investigated regarding the identification of reference genes suitable for the normalization of real-time quantitative PCR data. In this study, ten candidate reference genes (Act, CACs, EF-1 alpha, GAPDH, His3, PsaH, Sand, PP2A, ss-Tub and Ubq) were evaluated to determine the most stable internal reference for quantitative PCR normalization in cork oak. The transcript abundance of these genes was analysed in several tissues of cork oak, including leaves, reproduction cork, and periderm from branches at different developmental stages (1-, 2-, and 3-year old) or collected in different dates (active growth period versus dormancy). The three statistical methods (geNorm, NormFinder, and CV method) used in the evaluation of the most suitable combination of reference genes identified Act and CACs as the most stable candidates when all the samples were analysed together, while ss-Tub and PsaH showed the lowest expression stability. However, when different tissues, developmental stages, and collection dates were analysed separately, the reference genes exhibited some variation in their expression levels. In this study, and for the first time, we have identified and validated reference genes in cork oak that can be used for quantification of target gene expression in different tissues and experimental conditions and will be useful as a starting point for gene expression studies in other oaks.