The efficiency of five methods of conduction of segregating populations was
compared in common beans. For this experiment, the segregating population
from cross between the cultivars Carioca and Flor de Mayo was used. The met
hods pedigree, bulk, single seed descent (SSD), F-3 derived bulk and F-2 de
rived bulk were conducted as established for each method. They were evaluat
ed in two locations, Lavras, in the south of Minas Gerais, and in Patos de
Minas, located in the Alto Sao Francisco region. Utilizing a 18 x 18 triple
lattice design, 320 families were evaluated, with 64 being derived from ea
ch method, with the parents and two other checks. Estimatives were obtained
from genetic and phenotypic parameters for grain yield data (g/plots). The
main criteria utilized to compare the breeding methods were the average pe
rformance of the families, the expected response with different selection i
ntensities, and the number of families in each method with superior perform
ance to a particular standard. There were no significant differences betwee
n the methods to obtain the superior families, that is, if correctly conduc
ted, all the methods enable similar success with the selection. Considering
the estimatives of the genetic and phenotypic parameters, together with th
e ease and flexibility of conduction, the bulk and SSD method proved to be
the most advantageous.