The effectiveness of breeding for soybean ( Glycine max (L.) Merr.) hypocotyl length at 25 C was evaluated by examining segregation from crosses between long, intermediate, and short hypocotyl cultivars and performance of the parent cultivars. Progeny were evaluated in F 1 and F 2 as individual seedlings and in F 3 as the mean of 24 seedlings for each F 2 line. All tests for hypocotyl length were conducted in the dark in growth chambers maintained at 25 ± 1 C. Evaluation of individual seedlings for the parent cultivars indicated that there was an environmental influence on hypocotyl length, particularly for the intermediate cultivars. There was little overlap in the single plant distributions for long and short hypocotyl cultivars, but the distributions for intermediate cultivars had considerable overlap. Progeny from crosses between long hypocotyl cultivars were similar to the parents in hypocotyl length. Crosses of long ✕ intermediate and long ✕ short parents had segregation typical of a major gene with dominance for long hypocotyls. The F 2 lines in F 3 generally were within the range of the parents for intermediate ✕ intermediate, intermediate ✕ short, and short ✕ short crosses. Selection for genetic differences in hypocotyl length at 25 C is possible. To select for long hypocotyl genotypes, crosses should include at least one long hypocotyl parent. In crosses involving only intermediate and short hypocotyl parents, transgressive segregates may be found that exceed the length of the best parent; however, the frequency of superior segregates in such crosses probably will be low. Selection for hypocotyl length on a single plant basis is hampered by environmental effects. In a segregating population, selection for long hypocotyl genotypes on a single plant basis will eliminate many of the short hypocotyl genotypes, but will include some of the intermediate genotypes. A progeny test is the best way to evaluate a genotype's genetic potential for hypocotyl length.