<p>Haplotype-based genomic selection is considered a more efficient alternative for genomic selection under epistasis. We evaluated the efficacy of haplotype-based genomic selection under seven digenic epistasis types and the significance of including the additive x additive (AxA) effect in selection in pure line breeding. We simulated a genome composed of 10 chromosomes of 100 cM, including 1000 genes and 49,825 SNPs. We assumed positive dominance and 30% epistatic genes. We performed genomic selection using haplotypes of sizes 3, 6, and 9. We also processed low-cost genomic selection using 20% of the SNPs. We assessed selection efficacy using realized total genetic gain. We also evaluated the correlation between prediction accuracy and realized genetic gain and the decrease in genotypic variance over generations. Regardless of the epistasis type and SNP density, genomic selection was an efficient procedure for the development of superior pure lines. There was no effect of haplotype size on the efficacy of genomic selection. Under no selection, the AxA genetic value was negatively correlated with low magnitude ( &lt; −0.25) with the number of favorable genes. Under selection, the inclusion of the predicted AxA value for selection did not increase the selection efficacy. Haplotype- and SNP-based genomic selection processes were equally efficient. Compared with traits determined only by additive and dominance gene effects, epistasis can decrease the total genetic gain. Under epistasis, there was a highly positive correlation between genetic gain and prediction accuracy. There was no difference between genomic selection procedures in terms of the decrease in genotypic variance.</p>