A field trial was carried out on the Field Crops Research Station of State board of Agricultural Research Abu-Ghraib. Five inbred lines of maize were used in this study using full diallel cross in spring season in 2008 to produce twenty F1 crosses. Seed crosses were grown in a randomized complete block design with three replicates in fall season 2008. The objective of this study was to identify and select superior maize inbred lines based on their performance as well as to evaluate their single-cropss hybrid performance based on results of analysis of combining ability, gene action and estimation of some genetic parameters using the approach of Griffing (1956b) the third method with fixed model. The results of statistical analysis showed that there were highly significant differences among crosses for all the studied traits except the number of ears/ plant was not significant. Results of genetic analysis showed that the MSgca, MSsca, and MSrca, were high significant in all studied characters except for number of rows/ ear in (GCA) and for days to 50% silking and number of rows/ ear were not significant added for ear length was significant in (RCA). Indicating the importance of both additive and non additive gene action in the control of these characters. The inbred line AGR11 was found as the best general combiner for ear length (0.813), rows number (0.200), number of grains/ row(2.36) and grains yield/ plant(9.58). The best cross (AGR3 x AGR21) and (AGR11 x AGR21). These crosses showed good positive (SCA) effects along with better mean performance (40.9, 196.8) and (41.9, 171.8) respectively for grain yield per plant, and most of the traits. The ratio of (δ²gca) to the (δ²sca) and (δ²rca) was less than 1.0 in all studied characters, except for ear length, number of grains / row and grains yield which was more than 1.0 in reciprocal crosses. The values of dominance genetic variance were more than that additive genetic variance for all characters in diallel crosses, while the values of additive genetic variance were more than that dominance genetic variance except for the ear height, leaf area and number of rows per ear for all charscters in reciprocal crosses. This reflects the exceeded one for the value of average dominance degree for all studied characters in both diallel and reciprocal crosses except for days to 50% silking, ear length, number of grains/ row and grains yield/ plant in reciprocal crosses which was less than 1.0. As for broad sense heritability it was high for all studied characters in both diallel and reciprocal crosses except for days to 50% silking, number of rows/ ear and grains yield/ plant which were low in reciprocal crosses, but narrow sense heritability was low for all studied characters in diallel crosses while it was heigh for most of the traits in reciprocal crosses. The results indicate that some inbred lines could be used in a breeding program to develop new versions of high yield per plant and SCA to produce better grain yield hybrids, and most studied traits were under over dominance gene action. This showed that developing elite hybrids were the best method for improving maize grain yield.