Eight selected S₄ yellow inbred lines derived from the wide genetic base Population Gemmeiza Yellow maize at Mallawy Agricultural Research Station, were top-crossed with each of three commercial inbred line testers, namely, Gm-1001, Gm-1002 and Gm-1004 in 2005 season. The 24 top-crosses along with the check yellow maize SC-155 were evaluated at Sakha and Mallawy Agriculture Research Stations, ARC in 2006 growing season for grain yield (ard/fed), days to 50% silking, plant height and ear height. The obtained results revealed that mean squares due to top-crosses, lines, testers and lines × testers were highly significant for grain yield at each location.  The combined data for grain yield over the two locations showed highly significant variances for locations, crosses, lines, testers and interactions of lines x testers and location, crosses x locations and lines x locations and well as line x tester x location interaction. Two inbred lines (L-3 and L-5) had positive GCA effects and were good combiners for grain yield. The tester line Gm1002 was a good general combiner for both grain yield and days to 50% silking. Variance due to SCA was greater in magnitude than that due to GCA  for silking date and grain yield at Sakha and over locations and for silking date and ear height at Mallawy indicating that non-additive genetic variance was the major source of genetic variation. The magnitude of δ²GCA × Loc interaction was larger than δ² SCA × Locfor ear height and grain yield.  Results showed also that crossing the tester Gm-1004 with either of the inbred lines L-7 and L-5 produced the best two single crosses which significantly surpassed the check single cross 155 (20.74 ard/fad.) by 3.56 and 3.05 ard/fad., respectively.  However, from the combined data, the top crosses (L-8 x Gm-1002), (L-7 x Gm- 1002), (L-6 x Gm-1002), (L-4 x G-1004), (L-4 x Gm-1001), (L-3 x Gm-1001) and (L-2 x Gm-1004) produced 22.93, 22.65, 22.58, 22.26, 22.23, 22.44, and 22.35 ard/fad., respectively, relative to the commercial check SC. 155 (20.74 ard/fad.).  These nine yielded top crosses have to be advanced to the other steps for evaluation and testing.