Cotton aphids, Aphis gossypii Glover, were collected over a three-year period from cotton fields in Alabama and Texas and screened for resistance to four pyrethroid and eight organophosphate insecticides. These insecticides were also bioassayed against mummiform Lysiphlebus testaceives, an aphid parasitoid. Resistance ratios were based on comparisons of the LC(50), Values of field populations to the LC(50) values of a susceptible laboratory strain. During 1988, resistance to the pyrethroids bifenthrin and cypermethrin was low and moderate, and resistance to the organophosphates was variable.Resistance was highest (ca. 120-fold) to methamidophos. During 1988, resistance to multiple insecticides was found to increase rapidly within fields shortly after insecticide applications. Resistance during 1989 was almost nondetectable. This indicates that not only can resistance be developed quickly, but that it can also be lost quickly, suggesting induced resistance mechanisms may be involved. During 1990 resistance was widespread but occurred mostly at low levels. Resistance ratios and slopes of the probit regressions were variable within insecticides, and resistance was not always consistent within an insecticide class. This suggests that multiple mechanisms of resistance may be involved, and that resistance could be induced and is developed on a relatively small scale. All pyrethroids tested had similar activity against mummiform Lysiphlebus testaceipes while activity among the organophosphates was variable. The organophosphate dicrotophos was the most toxic insecticide tested while profenofos was least toxic. As a class, the pyrethroids were most selective toward aphids relative to the parasitoid. However, the organophosphates methamidophos, oxydemeton-methyl and profenofos were also relatively selective.