Apple (Malus × domestica Borkh.) is the most widely grown permanent fruit crop of temperate climates. Although commercial apple growing is based on a small number of globally spread cultivars, its diversity is much larger and there are estimates about the existence of more than 10,000 documented varieties. The varietal diversity can be described and determined based on phenotypic characters of the external and internal traits of fruit, which, however, can be modulated by environmental factors. Consequently, molecular methods have become an important alternative means for the characterisation of apple cultivar diversity. In order to use multilocus microsatellite data for determination of unidentified or misidentified apple varieties, a database with molecular genetic fingerprints of well-determined reference cultivars needs to be available. The objective of the present work was to establish such a database that could be applied for the molecular genetic determination of a large number of historic and modern, diploid and triploid apple cultivars. Based on the analysis of more than 1600 accessions of apple trees sampled in 37 public and private cultivar collections in different European countries at 14 variable microsatellite loci, a database with 600 molecular genetic profiles was finally obtained. The key criterion for considering a molecular genetic profile as confirmed and for including it into the reference database was that at least two accessions of the same cultivar of different provenances generated an identical result, which was achieved for 98% of the apple cultivars present in the database. For the remaining genotypes, the cultivar assignment was supported by a parentage analysis or by comparison to molecular genetic profiles available in published works. The database is composed of 574 scion cultivars, 24 rootstock genotypes and two species of crab apples. Of the 574 scion cultivars, 61% were derived from historic or old cultivars, many of which were grown in Central Europe in the past. The remaining scion cultivars are currently grown or available in testing programmes and may gain importance in the future. In order to validate the genotyping data, parentage analysis was performed involving cultivars and rootstocks that arose after 1900, for which information about at least one parent cultivar was available from pomological and scientific literature and the molecular genetic profiles of the assumed parent(s) were also present in our database. This analysis revealed the presence of null alleles at locus COL, however, when excluding this locus, a mean genotyping error rate of only 0.28% per locus was revealed, which points to a high reliability of the dataset. The datasets with 14 and 13 loci (excluding locus COL) showed a high degree of discrimination power, with a combined non-exclusion probability of identity of 2.6 × 10 -20 and 3.4 × 10 -19 . Five of the microsatellite loci analysed in the present study overlapped with another published dataset and after the application of conversion values, it was p...