Genomic prediction allows breeders to make selections based on the genomic estimated breeding value (GEBV) of selection candidates. GEBVs are assigned using a prediction model trained with genotypes and phenotypes from a training population. For this reason, the effectiveness of genomic selection is strongly tied to the prediction accuracy of the model used to estimate breeding values and the training population used to inform the model. The aim of this study was to evaluate the accuracy of the ridge regression best linear unbiased prediction (rrBLUP) model across different traits, parent population sizes, and breeding strategies when estimating breeding values in Phaseolus vulgaris. The model was trained on a simulated population genotyped for 1010 SNP markers including 38 known QTLs identified in the literature (Lin, 2022). Simulation results revealed that realized accuracies fluctuate depending on the factors investigated: trait genetic architecture, breeding strategy, and the number of initial parents involved in the breeding program. Trait architecture and breeding strategy appeared to have a larger impact on accuracy than the initial number of parents. Generally, maximum accuracies were achieved under a mass selection strategy followed by pedigree and single-seed descent methods. This study also investigated model updating, which involves re-training the prediction model with a more relevant set of genotypes and phenotypes. While it has been repeatedly shown that model updating generally improves prediction accuracy, it benefitted some breeding strategies more than others. For low heritability traits (e.g., yield) conventional phenotype-based selection methods showed consistent rates of genetic gain, but genetic gain under genomic selection reached a plateau in after fewer cycles.