The present investigation was carried out at the Agricultural Experiment station, Faculty of Agriculture, Alexandria University, during 1996,1997 and 1998 seasons. In this study two methods have been used for the early evaluating of inbred lines. These methods are S₁ per se and test cross performance. The Following results were obtained : The mean of S₁ was smaller than the test cross for all the characters under study. This reduction is function of inbreeding depression. It ranged between 5.8 for 100-kernel weight to 30% for grain yield . Highly significant difference were obtained among the two methods of selections for all character. Also the interaction between methods and S₁ lines were significant for all characters except ear width, moisture percent and days to mid-silking. This would indicate that testing of S₁ lines would be similar for either methods of evaluation for ear width, moisture percent and days to mid-silking only. Heritability estimates were in general higher for test cross evaluation. The heritability values for the combined evaluation was less than from single method. Magnitude of heritability would be reflected on the expected genetic gain from selection which was 12.6% for S₁ per-se and 15.1% for test cross. The interaction was highly significant for all characters except ear width, moisture % and days to mid-silking. The association between the two methods would be weak for the significant characters. The phenotypic correlation between the means of lines for both methods was significantly for grain yield , ear width, 100-kernel weight and moisture content. However their values were less than 0.5 indicating that the degree of association were weak. Only six lines were common in both methods, while 8 lines were common between S₁ and the combined evaluation and 12 lines between the test cross and the combined evaluation among the best 20% of lines. Therefore it was concluded that the two methods were different in the identification of superior lines.