Abstract Context Identifying species in the European white oak complex has been a long standing concern in taxonomy, evolution, forest research and management. Quercus petraea, Q. robur, Q. pubescens and Q. pyrenaica are part of this species complex in western temperate Europe and hybridize in mixed stands, challenging species identification. Aims Our aim was to identify diagnostic single nucleotide polymorphisms (SNPs) for each of the four species that are suitable for routine use and rapid diagnosis in research and applied forestry. Methods We first scanned existing whole-genome and target-capture data sets in a reduced number of samples (training set) to identify candidate diagnostic SNPs, ie genomic positions being characterized by a reference allele in one species and by the alternative allele in all other species. Allele frequencies of the candidates SNPs were then explored in a larger, range-wide sample of populations in each species (validation step). Results We found a subset of 38 SNPs (ten for Q. petraea , seven for Q. pubescens , nine for Q. pyrenaica and twelve for Q. robur ) that showed near-diagnostic features across their species distribution ranges with Q. pyrenaica and Q. pubescens exhibiting the highest and lowest diagnosticity, respectively. Conclusions We provide a new, efficient and reliable molecular tool for the identification of the species Q. petraea, Q. robur, Q. pubescens and Q. pyrenaica , which can be used as a routine tool in forest research and management. This study highlights the resolution offered by whole-genome sequencing data to design diagnostic marker sets for taxonomic assignment, even for species complexes with relatively low differentiation.