The objective of this study was to investigate the most efficient way to select F4:5 progenies derived from the Icatu and Catimor groups of Coffea arabica and to study the genotypic correlations between the traits related to coffee grain physical quality, diseases, and productivity. A combination of the predicted additive values for production capacity when considering seven harvests and a simultaneous selection for a high sieve percentage and resistance to specific diseases during the last harvest was used. Analyses that selected progenies without considering the genotype × harvest interaction provided inaccurate results, distorting the progeny ranking. Coffee leaf rust and brown eye spot were correlated, showing the possibility of simultaneously selecting for resistance to both diseases. Of the 68 progenies studied here, five showed satisfactory agronomic traits. Our findings will contribute substantially to the development of new coffee cultivars that will allow us to reduce pesticide use.
