Human activities including those in crop gene banks are subject to errors, especially during seed multiplication and maintenance of seed germination. Therefore, the most serious problem of gene banks is authenticity of the accessions and their genotypic purity. There are many methods for determining the identity of varieties, but comparisons between current data and past records are not easy since the latter are often missing. Breeding barley resistant to powdery mildew caused by Blumeria graminis f. sp. hordei (Bgh) was traditionally based on incorporating major genes into new varieties and the results have been published. Our goal was to identify resistance genes to powdery mildew in accessions of the Czech spring barley core collection and compare these data with earlier information to establish the authenticity of the accessions. Two hundred and twenty-three accessions of the collection including 665 single plant progenies were tested. Sixty-four selected reference isolates of Bgh representing the world diversity of the pathogen were used for resistance tests. Twenty-two known resistance genes were postulated either separately or in combinations. In the collection, 151 homogeneous accessions were found, but the resistances of nine of them were inconsistent with published data and in 12 accessions their authenticity is doubtful. The remaining 72 accessions were heterogeneous and comprised 176 resistance genotypes, 54 of which were probably mechanical admixtures of other varieties. There are several pathogens of cereals, e.g. rusts and mildews, against which many resistance genes in host crops have also been exploited. Knowledge of these resistances can assist in maintaining pure and genuine stocks in gene banks. Seed purity and the authenticity of accessions can subsequently be checked with more advanced methods.