The objective of this study was to evaluate genetic variability and select putative mutants of Terra Maranhão plantain cultivar (AAB genome) subjected to gamma radiation based on agronomic data and inter simple sequence repeat molecular marker profiles using the Ward-MLM strategy. A total of 233 irradiated plants and 41 controls were assessed. The agronomic and molecular data were subjected to the Ward-MLM statistical algorithm in the SAS program. Cluster analysis was performed by the average distance method (UPGMA), based on the distance matrix of the Gower algorithm, and the cophenetic correlation coefficient calculated using the R software. The distance between the putative mutants ranged from 0.321 to 0.524, with an average distance of 0.426, and a cophenetic correlation coefficient of 0.79. Three putative mutants, which were selected based on the best agronomic traits and low height, will undergo further evaluation in the next stages of the banana breeding program at Embrapa. These results describe the first attempt of using combined data of Terra Maranhão plantain cultivar for the purpose of selecting mutants and assessing genetic variability.