To study the expression of pathogenicity-related genes in Erwinia amylovora, seven candidate reference genes (ffh, glyA, gyrA, proC, pykA, recA, rpoB) were selected and validated with the following five different mathematic algorithms: geNorm, NormFinder, BestKeeper, the delta CT method and the RefFinder web-based tool. An overall comprehensive ranking output from each of the selected software programs revealed that proC and recA, followed by ffh and pykA, were the most stably expressed genes and can be recommended for the normalization of RT-qPCR data. A combination of the three reference genes, proC, recA and ffh, allowed for the accurate expression analysis of amsB and hrpN genes and the calculation of their fold change in E. amylovora after its infection of susceptible and resistant apple cultivars. To the best of our knowledge, this is the first study presenting a list of the most suitable reference genes for use in the relative quantification of target gene expression in E. amylovora in planta, selected on the basis of a multi-algorithm analysis.