Bacterial wilt is among the major diseases of tomatoes and is currently considered the most important in Brazil. The development of resistant cultivars is the strategy that presents greater efficiency and viability, being the mixed modeling approach allows the selection of genetic materials with major accuracy. This work defines strategies for the development and selection of segregating strains F2:3 from the estimation of genetic parameters and individual mean components using mixed models. A greenhouse experiment was conducted in a randomized complete block design with four replicates, in which the resistance of 43 progenies F2:3 segregants obtained from the F2 generation from the cross between the parents was evaluated, having as parents Yoshimatsu (Resistant) and IPA-7 (Susceptible). Components of variance and genetic parameters were estimated, followed by ranking by additive genetic values. The selection of F2:3 progenies resistant to bacterial wilt was more effective when the household information was used. Mixed model procedures for the selective accuracy of 77.98%, verified efficiency in the selection of F2:3 progenies resistant to bacterial wilt evaluated at 20 days after inoculation. A total of 240 resistant and moderately resistant individual progenies were selected with severity scores between 1 and 2.49. Families presented the new averages between 1.0 and 2.87, making 23 families F2:3 that will continue the tomato breeding program.
