In this paper we demonstrate a method for measuring the effect of spatial interactions on the adoption of hybrid rice using a unique, nationally-representative data set from Bangladesh. Based on results from a generalized spatial two-stage least squares estimation, we have shown that neighbor effects are a significant determinant of hybrid rice adoption. Further, using two specifications of spatial systems, we show that having a network including nearby hybrid rice adopters is more influential than having a network of more distant hybrid rice adopters, and that merely have a network with a large number of adopters may be relatively meaningless if they are far away.