A simultaneous equation generalized probit model is estimated to determine factors affecting technology adoption by California dairy farmers. Since productivity and technology choice are jointly determined, a single-equation approach to determine whether productivity affects technology adoption is subject to simultaneity bias. Since the system of equations contains both continuous and discrete endogenous variables, generalized probit is used. The findings indicate that the biased single-equation estimates tend to exaggerate relationships with explanatory variables, and in some cases, lead to different implications. This emphasizes the need to use the consistent and asymptotically more efficient generalized probit results to account for simultaneity.