The present study was conducted to assess the variation among 67 aggregatum onion accessions using 15 agro-morphological characters at the Department of Vegetable Science, Tamil Nadu Agricultural University, Coimbatore during rabi 2022–23. Based on per se performance, the populations Aca 15, Aca 9 and Aca 25 were found superior to the check variety CO (On) 5 for yield and yield-contributing traits. The results revealed that the first six principal components explained about 70.39 % of the total variation among the fifteen characters in 65 onion accessions, with eigenvalues greater than 1. PC1 accounted for 18.63 % of the total variability for all the traits. The traits such as number of leaves, number of shoots, polar and equatorial diameter, number of bulbs, weight of bulbs, total soluble solids and pest and disease incidence contributed the most to the observed variability. Principal component analysis (PCA) will help researchers understand, utilize, conserve and manage the collection for a more efficient contribution to onion research and cultivation. Hence, a plant breeder must place special emphasis on these traits while conducting selection in a crop improvement programme in onion.