An exotic Zea mays L. population ('Tuxpeno') was adapted to North Carolina conditions by first introducing genes for adaptability from two North Carolina varieties ([('Jarvis' X 'Indian Chief')'Tuxpeno']'Tuxpeno') including four generations of intermating, and then selecting for adaptability using maturity as the primary measure. The study evaluated selection for adaptability and the diversity available between adapted 'Tuxpeno' and the local varieties, 'Jarvis' and 'Indian Chief'. Analytical procedures were developed to quantify the diversity between populations and the complementation of local varieties by introduced germ plasms. The analyses utilized the specific effects available from the diallel mating design.Three replicate selections responded similarly under simple recurrent mass selection (1/10) for the earliest disease-free plants initially and additionally for plant types (primarily height) in the final generation. The 1/4 local germ plasm permitted rapid adaptation of 'Tuxpeno' gene pool to local conditions. The adapted 'Tuxpeno' populations yielded similarly to the local populations with an average heterosis for grain yield of 28% when crossed to the local populations used as source of genes for adaptability. The diversity found between adapted 'Tuxpeno' lines and these local varieties based on genes affecting grain yield was 1.5 to 2.5 times that measured between the local varieties ('Jarvis' and 'Indian Chief'). Diversity lost through intergradation with local material was a reasonable investment. Yield genes introduced from 'Tuxpeno' complemented local gene pools through nonadditive, primarily dominance-associated, gene effects. Reassortment of major gene blocks apparently occurred leading to significant divergence among replicate selections involving both additive-associated and dominance-associated gene effects.