The zucchini (Cucurbita pepo) is an important food crop, the transcriptomics of which are a fundamental tool to accelerate the development of new varieties by breeders. However, the suitability of reference genes for data normalization in zucchini has not yet been studied. The aim of this study was to assess the suitability of 13 genes for their potential use as reference genes in quantitative real-time PCR. Assays were performed on 34 cDNA samples representing plants under different stresses and at different developmental stages. The application of geNorm and NormFinder software revealed that the use of a combination of UFP, EF-1A, RPL36aA, PP2A, and CAC genes for the different experimental sets was the best strategy for reliable normalization. In contrast, 18S rRNA and TUA were less stable and unsuitable for use as internal controls. These results provide the possibility to allow more accurate use of qPCR in this horticultural crop.