Exploration of genetic variation may help to select diverse parents for open pollinated variety and hybrid development program in onion (Allium cepa L.). Thirty-five onion genotypes were analyzed by two-way cluster analysis and principal component analysis (PCA) on the basis of phenotypic traits. These genotypes were grouped in five different clusters. PCA provided information for variation in genotypes on the basis of studied traits. Ten genotypes were selected from five diverse clusters and hybridized in a line × tester mating fashion. Twenty-four F1 combinations were developed and raised along-with parents for assessment of hybrid vigor, commercial heterosis and genetic correlation. Data was analyzed for morphological and biochemical traits. Maximum commercial heterosis (heterosis computed over commercial standard/check variety) for bulb yield per plot was observed in combination MKS-SGB × MKS-GBP-01, MKS-57404 × Super Sarhad and MKS-SGB × Mustang (55.77%, 36.70% and 32.90%), respectively. Maximum better parent heterosis for bulb yield per plot was in combinations MKS-SGB × MKS-GBP-01 (31.01%), followed by MKS-TEG × Mustang, MKS-57404 × Super Sarhad and MKS-SGB × Mustang (13.76%, 12.49% and 11.79%), respectively. Genetic correlation coefficient showed that bulb yield per plot showed highly significant and positive association with single bulb weight, bulb diameter, significant positive association with leaf density (number of leaves per plant), leaf width (diameter) and significant but negative association with total soluble solids. Yield per plot showed a negative association with days to maturity. Present study revealed that cluster analysis and PCA can play an important role in effective parent selection without compromising genetic diversity for development of potential hybrids for commercial exploitation.