Despite the importance of location-specific adaptive crop breeding research, past reforms of breeding systems in Nigeria have focused more on centralizing the breeding activities into fewer locations. This has been based partly on the premise that such research systems can still effectively meet the need for a diverse set of varietal technologies that are suitable for different agroecological conditions through the use of numerous outstations and multilocational trials, regardless of the locations of the headquarters or the outstations where breeders are located. However, little empirical evidence exists to support this premise. Using panel data for agricultural households in northern Nigeria, as well as spatial data on agroecological factors, this study fills this knowledge gap. Specifically, it empirically shows that agricultural productivity and technical efficiency at farm household level is significantly and positively affected by similarity between the agroecological conditions of the locations of these households and where major crop breeding institutes are headquartered in Nigeria, namely Maiduguri, Kano, Zaria, Badeggi, Ibadan, and Umudike, after controlling for the agroecological conditions and various relevant household characteristics of these households. These findings suggest that where improved varieties are developed or evaluated affects agricultural productivity and technical efficiency in different locations. Overall agricultural productivity in Nigeria can be significantly increased not simply by increasing support for public sector varietal development, but by doing so in a manner that increases the similarity in agroecological conditions between areas where crop breeding is conducted and the areas where farm households produce those crops.