Explaining the evolution of evolvability is, for many, a central problem in evolutionary biology. In this paper I offer a solution to at least an aspect of this problem by giving an account of the evolution of one recognised source of evolvability—"exploratory behaviour" in development. In particular, I make use of the similarities between development systems that exhibit "exploratory behaviour" and learning systems in order to make some claims about the enabling conditions for the evolution of exploratory behaviour.