Breeders attempt to conduct selection under environmental conditions representative of the target environment. In the U.S. southern Great Plains, more than 50% of the wheat area may be used for the dual purpose of producing forage for cattle grazing and harvesting grain, but breeding of cultivars likely occurred in an environment managed for grain production. We tested the hypothesis that genetic improvements accrued over time in agronomic performance may be compromised, or be differentially expressed, in a forage-plus-grain system compared to the grain-only system under which improvement was initially targeted. Two field experiments were conducted in each of three years, employing management components appropriate to each system, and using a historical set of 12 cultivars chosen for their widespread adoption in the region. Substantial genetic improvement has occurred in hard red winter wheat yield, without adverse effects on test weight or grain protein content. The magnitude of genetic gain was higher in the grain-only management system under which these cultivars were originally selected. In two of the three years, we found no significant trend for improvement in grain yield under the dual-purpose system, prompting an expansion of our breeding objectives to incorporate selection pressure for dual-purpose adaptation.