Breeding for adaptation to adverse environment should capitalize on additive genetic variance while simultaneously reducing genotype by environment interactions. In the present work, 400 S1 maize lines from a population resistent to drought from the Collaborative Regional Maize Breeding Project were evaluated under limited moisture conditions. These lines were evaluated in a lattice design of 20 x 20 with two replications per location in three different moisture environments. Selection pressure was 10%. Selection differentials for the selected fraction for recombination and experimental varieties were high. The selection differential was higher under limited moisture conditions confirming highly significant differences observed in the analysis of variance for lines. The experimental variety and the selected fraction had yield reductions with drought of 36%, in comparison to the population bulk which was reduced by 51 % with drought, demonstrating the efficacy of applied selection pressure. Within the variables studied, ears per plant correlated the highest with yield under limited moisture conditions. The anthesis-silking interval did not correlate as high with yield (interval of.8 to 5.9 days), possibly because the population had becn improved for this eharaetcristie. Estimatcd stability paramcters (β = 1 r = 0.9) identified the selected lines for synthctie varictics as stable cntries with a eonsistent response aeross the 3 moisture environments. Broad-sense heritability estimates (H2 = 0.43 and H2 = 0.64**) for yield and ears per plant under drought indieate that oneean expeet to obtain substantial selection progrcss by increasing the frequencies of favorable alleles which condition the hcritability of these characteristics.