Variations in fiber yield among cultivars and strains of upland cotton have been shown to be largely dependent upon genetic factors with additive effects. Heterosis has been demonstrated for some crosses, and this trait is thought to be dependent mostly upon dominance effects, although there is some evidence for epistatic effects. Generally, cotton displays about the same proportion of general and specific combining effects for desired product yield as several other crops, including corn and sorghum, species that have been greatly improved through the use of hybrid cultivars.

Experiments on hybridization performed in the past showed occasionally high levels of PUH (potentially useful heterosis), that is to say, hybrids that yielded significantly more fiber than the better parent. El-Adl and Miller selected within a family stemming from one such hybrid and found that three cycles of recurrent selection produced inbred lines that exceeded the yield of the heterotic Fl used to initiate the cycle. Moreover, the recovered lines, when intercrossed, yielded hybrids that displayed modest amounts of PUH.

Topcrosses involving commercial cultivars and various breeding lines have shown that the highest yielding hybrids displaying PUH stemmed from certain of the best yielding parents. Thus as cotton is improved by conventional selection any expectation of producing superior hybrids would, seemingly, stem from the crossing of the resulting cultivars.

Meredith and Bridge have shown that cotton cultivars offered over the previous 14 years have advanced about 11.4 kg/ha/yr in potential fiber yield which, averaged out to about 0.84% gain per annum. Heterosis in such cotton is reckoned to be a thin line along the top of such ongoing progress.