using two diverse approaches, GBLUP and BayesC, using simulated data under different quantitative trait locus (QTL) effect distributions. Data were simulated with three different distributions for the QTL effect which were uniform, normal and gamma (1.66, 0.4). The number of QTL was assumed to be either 5, 10 or 20. In total, 9 different scenarios were generated to compare the markers estimated breeding values obtained from these scenarios using t-tests. In comparisons between GBLUP and BayesC within different scenarios for a trait of interest, the genomic estimated breeding values produced and the true breeding values in a training set were highly correlated (r>0.80), despite diverse assumptions and distributions. BayesC produced more accurate estimations than GBLUP in most simulated traits. In all scenarios, GBLUP had a consistently high accuracy independent of different distributions of QTL effects and at all numbers of QTL. BayesC produced estimates with higher accuracies in traits influenced by a low number of QTL and with gamma QTL effects distribution. In conclusion, GBLUP and BayesC had persistent high accuracies in all scenarios, al-though BayesC performed better in traits with low numbers of QTL and a Gamma effect distribution.