This study used state-level and sub-state, region-level data to econometrically estimate factors explaining regional differences in the speed an extent of the diffusion of Bt cotton varieties from 1996 to 1999. Differences in diffusion patterns could largely be explained by supply-side factors (availability of Bt varieties adapted to local production conditions) and demand- side factors (such as expected pest control cost savings, expected yield gains, market prices and program payment rates). The study next estimated the effect of Bt cotton adoption on the change in regional insecticide use from the pre-Bt 1991-5 average, controlling for other factors affecting insecticide use. Bt adoption had a large and statistically significant negative impact on insecticide use. Bt cotton adoption contributed to a drop in insecticide use of 0.71 applications per total U.S. cotton acres in 1996 and a drop of 1.68 applications in 1999.