In agriculture and plant breeding, plant traits may be favoured because they benefit neighbouring plants and ultimately increase total crop yield. This idea of promoting cooperation among crop plants has existed almost as long as W.D. Hamiltons inclusive fitness (kin selection) theory, the leading framework for explaining cooperation in biology. However, kin-selection thinking has not been adequately applied to the idea of cooperative crops. Here, I give an overview of modern kin selection theory and consider how it explains three key strategies for designing cooperative crops: (1) selection for a less-competitive plant type (a communal ideotype); (2) group-level selection for yield; and (3) exploiting naturally-selected cooperation. The first two strategies, using artificial selection, have been successful in the past but suffer from limitations that could hinder future progress. Instead, I propose an alternative strategy and a new colonial ideotype that exploits past natural selection for cooperation among the modules (e.g., branches or stems) of individual plants. More generally, I suggest that Hamiltonian agriculturea kin-selection view of agriculture and plant breedingtransforms our understanding of how to improve crops of the future.