The impact of selection in irrigated environments on improving seed yi
eld in drought environments has not been extensively researched in soy
bean [Glycine max (L.) Merr.]. Our objectives were to evaluate cultiva
r x irrigation interaction, its impact on selection for yield in droug
ht environments, and drought tolerance in soybean: drought tolerance b
eing defined as better relative yield ranking in drought environments
than in nondrought environments. Seven data sets were obtained from Ar
kansas trials of Maturity Group V cultivars. Each contained data from
irrigated and nonirrigated trials from one location from two consecuti
ve years with drought, Significance of genotype x irrigation interacti
on was determined for each set along with heritability of yield in bot
h irrigation treatments and the genetic correlations of irrigated and
nonirrigated yield from one gear with nonirrigated yield in the other
year, Rank correlations and crossover interactions between irrigation
treatments were analyzed. The effect of selection in both irrigation t
reatments on yield in drought environments was simulated, Genotype x i
rrigation interaction was significant in six data sets. Entry-mean her
itability was higher in the irrigated trials (0.74) than in nonirrigat
ed trials (0.37), Rank changes between irrigated and nonirrigated trea
tments rarely consisted of large yield differences. The genetic correl
ations of yield from irrigated and nonirrigated trials from one year w
ith yield From nonirrigated trials in another year were nearly equal.
The results indicated that selection from irrigated trials would impro
ve yield in drought environments better than selection from nonirrigat
ed trials. Drought tolerant reactions occurred in only 16% of the two-
cultivar comparisons. While several cultivars appeared tolerant, their
yield in drought environments was not superior to several less tolera
nt cultivars. Nonirrigated trials may be useful in identifying a few u
nadapted cultivars.