This study was undertaken to estimate the general and specific combining ability of different female and male lines in F1 combinations for grain yield and some important traits and estimate heterosis as a criterion for developing superior hybrids. The twenty five hybrids and their parents were evaluated at eight environments; two planting dates at two locations in 2007 and 2008 seasons. The obtained data showed that highly significant differences among parents, F1 crosses and parents versus crosses over the two years and over all environments for traits. An evaluation of general combining ability variance components as estimated from male and/or female overall environments were larger than those of specific combining ability for days to 50% blooming, plant height and 1000-grain weight. However, opposite results were obtained for grain yield Mg ha-1. The estimates of general combining ability and specific combining ability variance components varied greatly from location to location and from early to late planting for days to 50% blooming, plant height and 1000-grain weight. Some parents having significant negative general combining ability for days to heading and significant positive for plant height, 1000-grain weight and grain yield were considered as good combiners. Significant positive heterosis in grain yield heterosis was found for more than half of the hybrids studied. Several cross combinations showed significant positive 1000-grain weight heterosis, significant negative days to heading heterosis and good performance. Since selection of grain sorghum hybrids in this study should be based on high grain yield, early maturing, taller plants and heavier grain weight. Information on general and specific combining ability and heterosis for those four traits could contribute to more efficient breeding program.