
This study aimed to determine the dissimilarity among landrace common bean genotypes through multivariate techniques aiming the selection of individuals to compose the crossing blocks and determine among the evaluated genotypes, which ones have proximity to an agronomic ideotype. The experiment was carried out during the 2014/2015 agricultural year in TenentePortela– RS, southern Brazil. The augmented blocks design was composed by the control treatments: BRS Estilo, IPR Uirapuru, IPR Tangara, BRSMG Realce, IAC Formoso, IPR Imperado, and IPRTiziu, and the regular treatments were constituted by the genotypes: LMGPP01, LMGPP2, LMGPP3, LMGPP4, LMGPP5, LMGPP6, LMGPP7, LMGPP8, LMGPP9, LMGPP10, LMGPP11, PORTELA16, PORTELA37, PORTELA38, PORTELA39, PORTELA64, and PORTELA65. The genotypes LMGPP09 and PORTELA16 are the most dissimilar, while LMGPP01 and LMGPP02 are the most similar ones. The BRSMG REALCE and PORTELA16 genotypes can be used as genitors in combination with other genotypes in the formation of segregating populations with wide genetic variability. The hierarchical UPGMA clustering method, the Tocher optimizing method, and the principal component analysis are consistent and complementary, enabling a reliable response in the data interpretation on the genetic dissimilarity in common bean.