The genetic variability of thirty-six sesame accessions were evaluated using molecular and morpho-agronomic data, aiming to identify divergent genotypes for further use in breeding program. Ten SSR markers and twenty seven morpho-agronomic traits were used to estimate genetic divergence by means of multivariate Tocher and UPGMA methods. The GENES program was used for statistical analysis of data. We observed that molecular and morpho-agronomic data were efficient to estimate divergence among the accessions. In molecular aspect, the ZM_22, ZM_45 and ZM_34 markers showed broad contribution by presenting polymorphism rates of 0.53, 0.44 and 0.39, respectively. The groups formed in the Tocher model were not similar to those formed by the UPGMA method. Among the groups formed, the study of genetic diversity allowed identification of characteristics of interest such as precocity in the genotypes ABG 591 and ABG 649. The degree of similarity assembling revealed the most similar (ABG 591, ABG 616, ABG 649, ABG 141 and ABG 688) and identifying the most divergent (ABG 648 and ABG 200) genotypes. These results revealed significant genetic variability among the investigated accessions of the sesame ABG that can be applied in the breeding programs.