Fifty inbred lines of maize acquired from various sources, were screened in the wirehouse for seedling traits under both normal and water deficit conditions. Six inbred lines were earmarked on the basis of various agronomic and physiological traits under water deficit condition. The inbred lines were sown in the field for making all possible crosses in diallel mating fashion. The F 1 crosses and their reciprocals alongwith the parents were sown in the field under normal and water stress environment using Randomized Complete Block Design in three replications. Normal irrigations were applied to one set of experiment, whereas 50% of the normal irrigation was applied to water stress experiment. Data for various morpho- physiological characters were recorded at different growth stages of the crop and then subjected to statistical analysis. Co-efficient of variability was found to be greater under water stress than under normal condition for majority of the seedling traits. Significant mean squares for all the parameters under normal and moisture deficit conditions depicted the presence of considerable genetic variability. Scaling tests were performed to check the adequacy of the data for analyzing additive-dominance model. The results showed that additive-dominance model was fully adequate for the traits like kernels per row, 100-kernel weight, grain yield per plant, cell membrane thermostability, stomatal conductance and canopy temperature under normal condition while plant height, ear height, days to silking, kernels per ear, 100-kernel weight, grain yield per plant, leaf temperature and canopy temperature depression under water stress condition. The data were partially adequate for the traits like ears per plant, days to tasselling, anthesis-silking interval, kernels per row, cell membrane thermostability and stomatal conductance under moisture deficit condition. Additive gene action for number of days to silking under normal water condition changed to non-additive gene action under water deficit condition. All other traits exhibited additive gene action under both conditions. Heritability estimates for yield related traits revealed maximum ability to transfer the desirable genes to the next generation. Estimation of components of variation exhibited greater estimates for GCA variance ( б 2 g) than SCA variance ( б 2 s) for majority of the traits under both conditions depicting the predominant role of additive genetic component except for days to silking and number of kernels per row under water deficit condition which displayed greater SCA variance ( б 2 s) than GCA variance ( б 2 g). Under water stress condition, the best performing crosses were NCIL-20-20 x D-109, NCIL-20-20 x OH-8 and D-114 x NCIL-20-20 and their reciprocals. Inbred lines NCIL-20-20, D-157 and D-114 proved to be high yielder parents under both normal and moisture deficit conditions. The information regarding results obtained during the current study may be used to evolve better parental inbred lines for developing various cross combinations which will be helpfull in maintaining yield sustainability in water deficit areas.