Long-term isolation of populations and interbreeding can lead to morphometric variations among fish populations. This study was conducted with the objective to delineate stock structure of Sperata aor on the basis of morphometric characters, using truss network of the fish body. In total, 407 S. aor samples were collected from the four sampling locations of the Ganga River, viz. Narora, Kanpur, Varanasi and Bhagalpur. In total, 28 truss distance measurements were created by joining 13 morphometric landmarks on the fish body. MANCOVA showed significant (P<0.001) morphological variance among the sampling locations for the target fish species. Univariate ANOVA showed significant (P<0.001) differences in each of the morphometric measurements among the fish from different sampling locations. Wilk’s lambda test of canonical discriminant function analysis showed significant (P<0.001) differences in morphometric measurements of the fish from all sampling locations. Discriminant function analysis using Jackknife (leave-one-out) cross-validation classification showed 87.5% correct classification of the individuals into their original populations. Mantel test showed an overall good correlation between the genetic and morphometric datasets in the selected fish species. The results of this study can be employed in formulating stock-specific management strategies for S. aor from River Ganga.