The aim of our study was to create a high-quality Holstein cow genome reference sequence and describe the different types of variations in this genome compared to the reference Hereford breed. We generated one fragment and three mate-paired libraries from genomic DNA. Raw files were mapped and paired to the reference cow (Bos taurus) genome assemblies bosTau6/UMD_3.1. BioScope (v1.3) software was used for mapping and variant analysis. Initial sequencing resulted in 2,842,744,008 of 50-bp reads. Average mapping efficiency was 78.4 % and altogether 2,168,425,497 reads and 98,022,357,422 bp were successfully mapped, resulting in 36.7X coverage. Tertiary analysis found 5,923,230 SNPs in the bovine genome, of which 3,833,249 were heterozygous and 2,089,981 were homozygous variants. Annotation revealed that 4,241,000 of all discovered SNPs were annotated in the dbSNP database and 1,682,230 SNPs were considered as novel. Large indel variations accounted for 48,537,190 bp of the entire genome and there were 138,504 of them. The largest deletion was 18,594 bp and the largest insertion was 13,498 bp. Another group of variants, small indels (n = 458,061), accounted for the total variation of 1,839,872 nucleotides in the genome. Only 92,115 small indels were listed in the dbSNP and therefore 365,946 small indels were novel. Finally, we identified 1,876 inversions in the bovine genome. In conclusion, this is another description of the Holstein cow genome and, similar to previous studies, we found a large amount of novel variations. Better knowledge of these variations could explain significant phenotypic differences (e.g., health, production, reproduction) between different breeds.