Background. This study aimed to explore the morphological traits associated with seed yield in garden cress and to recognize the direct and indirect effects of agronomic characteristics. Analyzing the relationships between traits and seed yield with understanding their causal effects helps breeders choose the best combination of yield components to improve economic performance. Materials and methods. A total of 64 accessions, including 60 international and 4 local Iranian genotypes, were evaluated across two locations (Ardabil and Eyvanki) using a randomized block scheme. Results. Biomass consistently emerged as the most influential trait across both environments, followed by leaf- and pod-related characters. Categorizing traits into primary, secondary, tertiary, and quaternary orders helped reduce multicollinearity and clarified the structure of trait interrelationships. Path analysis indicated several optimal selection pathways, such as LB → PP → PB → BIO → SY in Ardabil, and FLO → LW → LL → SY in Eyvanki. The bootstrapped estimates confirmed the stability and reliability of the identified trait effects. Conclusion. Finally, indirect selection based on biomass, leaf traits, and yield components can improve seed yield in garden cress. Also, simultaneous improvement of seed and leaf yield is feasible and should be considered in future breeding programs
