Relatively rapid and cost-effective genotyping using the OvineSNP50 chip holds great promise for the South African sheep industry and research partners. However, SNP ascertainment bias may influence inferences from the genotyping results of South African sheep breeds. Therefore, samples from Dorper, Namaqua Afrikaner (NA), South African Merino (SA Merino) and South African Mutton Merino (SAMM) were genotyped to determine the utility of the OvineSNP50 chip for these important South African sheep breeds. After quality control measures had been implemented, 85 SA Merino, 20 Dorper, 20 NA and 19 SAMM samples remained, with an average call rate of 99.72%. A total of 49 517 (91.30%) SNPs on the chip met quality control measures and were included in downstream analyses. The NA had the fewest polymorphic loci, 69.20%, while the SAMM, Dorper and SA Merino had between 81.16% and 86.85% polymorphic loci. Most loci of the SA Merino, Dorper and SAMM had a MAF greater than or equal to 0.3. In contrast, the NA exhibited a large number of rare alleles (MAF < 0.1) and a uniform distribution of other loci across the MAF range (0.1 < MAF ≤ 0.5). The NA exhibited the least genetic diversity and had the greatest inbreeding coefficient among the four breeds. The results of the Dorper, SA Merino, and SAMM compare favourably with those of international breeds and thus demonstrate the utility of the OvineSNP50 chip for these breeds. Effects of SNP ascertainment bias, however, could be seen in the number of non-polymorphic loci and MAF distribution of the three commercial breeds in comparison with those of the NA. The implementation of methods to reduce the effect of SNP ascertainment bias and to ensure unbiased interpretation of genotype results should therefore be considered for future studies using OvineSNP50 chip genotype results.