Diffusion of new knowledge and technologies in agriculture can offset the sometimes explosive nature of food price increases, especially in developing countries. Closely following the notion of innovative geographic clusters, this thesis examines knowledge flows in the US agriculture industry for evidence of innovative agglomeration. Each agricultural patent granted from 1972 - 2002 was spatially tagged using Geographic Information Systems software. The data indicate that a closer distance between any two patent origins increases the probability that one cites the other as prior art, with subtle interregional variations on the degree to which proximity advances agricultural innovation. Policies that exploit the relative ease of knowledge within localized networks can encourage development of cost-cutting technologies which can in turn lower world food prices.