The experiment was carried out at the field of Field Crop - College of Agriculture-Univ. of Baghdad, during spring and fall seasons of 2013, under two levels of Nitrogen; 200 and 400 kg N.ha-1, by using Randomized Complete Block Design, with four replications. The objectives of the experiment were to estimate the genotypic, phenotypic and environmental correlations and path coefficient of yield and its components that correlated with yield to rely on them as selection criteria for high yield of maize. Results of statistical analysis showed that studied traits contributed with 99% and 85% of yield variance at spring and fall under 200 kg N.ha-1. respectively, and 99% and 93% of yield variance at spring and fall under 400 kg N.ha-1 respectively. The trait number of grains plant-1 exhibited positive and highly significant genotypic and phenotypic correlation with yield, followed by other traits, while the grain weight correlated negative and highly significant correlation with yield , at spring under 200 kg N.ha-1. Traits e.g. number of grains row-1, number of grains Plant-1, weight of grain, and crop growth rate exhibited positive and highly significant correlation with yield. Under 400 kg N.ha-1. In both seasons, all traits had positive high significant phenotypic and genotypic correlation, except days to maturity was negative and high significant with yield. In spring and fall under 200 kg N.ha-1. grains plant-1 exhibited positive high value of direct effect in yield followed by grain weight . Crop growth rate exhibited positive direct effect in yield in fall season. Under 400 kg N.ha-1, in spring the highest direct effect was to grains plant-1 followed by other traits except grain row-1 was negative, while in fall the highest direct effect were for grains plant-1, grain weight and crop growth rate. It can be concluded that these traits could be used as selection criteria for yield improvement in maize in breeding programs.