The okra- and normal-leaf isolines from two cotton (Gossypium hirsutum L.) genotypes were grown on three soil types in the Mississippi Delta near Stoneville, MS from 1989 to 1996. There were 15 tests for DES 24-8ne and seven tests for MD 51ne. Soil types included a Bosket fine sandy loam, a Dundee silty clay loam, and a Beulah fine sandy loam. For the DES 24-8ne background, yields of okra-leaf were 5 to 41% greater than normal-leaf types (e.g. 1060 vs 1010 lbs/acre and 554 vs 393 lbs/acre) when the site was a Bosket fine sandy loam. Similar results were found for MD 51ne. In contrast to the Bosket site, yields of DES 24-8ne and MD 51ne normal-leaf were either equal to or greater than okra-leaf on the other two sites. Environments (site-years) were ranked based on yield of the normal-leaf isoline with 1 equal to the lowest yield. The percentage of okra-leaf advantage (or disadvantage) was plotted against these rankings. For DES 24-8ne and MD 51ne the correlations were negative (r= - 0.68*, n=15 and r= - 0.62 ns, n=7, respectively). The results indicate that an okra-leaf advantage may have occurred not because it exhibits a high yield potential, but because the yield of the normal-leaf was depressed by an unidentified factor. Among the suspected factors are excess N, excess water, and high tarnished plant bug (Lygus lineolaris) populations, (two of which are naturally occurring) that are common to several Mississippi Delta cotton growing sites. In such environments, okra-leaf appears to have a relative advantage over normal-leaf in at least two isogenic lines.