 This research sought to use genomic prediction to identify superior genotypes of jucazinho, uterpiodullus, a popular fruit tree species in Brazil. Genomic prediction was applied to a dataset consisting of 275 genotypes from Rio Novo do Sul, Brazil. Two different models were used, multi-trade, G-blup MT, and single-trade, G-blup ST. Both models showed similar predictive abilities, but the single-trade model produced greater selection gains than the multi-trade model. Based on these results, the best six genotypes were selected using the single-trade model. These genotypes have potential to be used in the development of superior genetic materials for the production of seedlings and the establishment of productive orchards. This article was authored by Guilherme Bravam Canal, Cynthia Aparecida, Valiotti Barreto, Francine Alves Noguera de Almeida, and others.