A series of laboratory and field experiments were initiated in 1995 to determine if laboratory assays with different Bacillus thuringiensis (Bt) commercial products accurately predict field efficacy against tobacco budworm (Heliothis virescens) on cotton. Concentration-mortality studies were conducted with H. virescens, cotton bollworm (Helicoverpa zea), and beet armyworm (Spodoptera exigua)in diet incorporation, leaf disc, and terminal assays with 11 different commercial Bt's. Results were compared to field studies when 5 rates of each Bt were tested against high population densities of H. virescens and H. zea. Data from the field studies failed to show differences in Bt products or rates. As a result, LC50's from the different assays were converted to H. virescens units, H. zea units, and S. exigua units and compared to a spray table study using terminals with H. zea. When data were converted to H. zea units most of the assay procedures correlated with the spray table data. This illustrates the importance of establishing Bt potency and the importance of using the target species as the assay orginism. With cotton this species should be H. virescens, but more data is required.