ABSTRACT This study provides a morphological and productivity analysis of 20 soybean (Glycine max L. Merr.) genotypes maintained in the Azerbaijani Genetic Resources collection (AZGR), with emphasis on the relationships between plant development characteristics and yield components. The 20 genotypes originated from Azerbaijan, Uzbekistan, Turkey, Canada, and Australia and were evaluated at the Saray Support Point, Institute of Genetic Resources, during 2024–2025. A hierarchical cluster analysis was performed using PAST v4.16 software (Hammer et al., 2001) based on predetermined morphological and yield-related criteria. Three distinct genetic clusters were identified. Cluster I (9 genotypes, including Kanata and Gen-8) was distinguished by the longest pod length and highest pod number per plant. Cluster II comprised predominantly Uzbek genotypes, further divided into two sub-clusters. Cluster III included genotypes with intermediate morphological profiles most similar to the Australian (Anjelica) and Kyota accessions in productivity metrics. The results demonstrate that genotype origin and genetic background significantly influence key yield components and provide a practical basis for targeted selection in soybean breeding programmes aimed at improving productivity under Azerbaijani agro-ecological conditions