Eradication of the boll weevil from the southeastern USA has prompted growers to consider cotton as a second crop after wheat. Although current acreage is minor, several thousand acres at most, the changing economics of soybean production makes cotton an attractive option to replace or supplement soybean in a wheat-soybean doublecrop production system. Available cotton cultivars for the southeast have been developed for full-season or early maturity in full-season production systems and generally not for late-planting in a doublecrop system. The issue in a cotton breeding program is whether resources should be directed at breeding for both production systems. Ideally, selection for yield and fiber properties in the conventional production system would lead to improved genotypes for both production systems. The objectives of this study were to evaluate the genotype x planting date interaction for lint yield and fiber properties using random cotton lines. Twenty-five unselected Pee Dee cotton lines were evaluated in conventional (May 1) and late-planted (June 7) production systems for two years at Florence, SC. Lint yield, fiber, and spinning properties were determined. Significant (p<0.05) genotypic variation existed for all of the fiber traits, yarn strength, and lint yield. There was no significant genotype x planting date interaction for any traits. The genetic correlation between planting dates as well as heritability of traits in each planting date were used to calculate predicted direct and indirect response to selection. Selection for lint yield, fiber, and spinning properties in the conventional environment resulted in progress in both production systems. Overall, initial selection in the conventional production system would be a sufficient breeding scheme to improve full-season as well as late-planted performance.