The genetic architecture of dark cutting was investigated with a case-control genome-wide association study on two groups of beef cattle analyzed separately and together (combined group). Groups I (n = 64) and II (n = 150) were genotyped using the 70K GeneSeek Genomic Profiler for Beef Cattle-HD and the 50K Illumina BovineSNP50v2 BeadChip, respectively. Dark cutting was analyzed as a binary trait (case versus control) using logistic regression in an additive model implemented in PLINK version 1.9. Significant loci were not identified when correcting for multiple testing (false discovery rate), suggesting that the trait is not controlled by genes with big effects, or the sample size was not large enough to detect these major genes. Regions harbouring single-nucleotide polymorphisms (SNPs) with a raw p < 0.01 using 1 MB window were analyzed for gene function using the ingenuity pathway analysis. For groups I, II, and the combined group, 449, 301, and 191 SNPs were identified, respectively. Genes identified were involved in pyruvic acid modification and release, 2-deoxyglucose clearance and disposal, sucrose recognition, energy production, and metabolism of carbohydrate. Although detected SNP associations require validation in a large population, results suggested the possibility for marker-assisted or genomic selection of beef cattle to reduce dark cutting.