This paper explores the idea that a competitive seed system may systematically underserve farmers in small, agro-ecological niches, leaving potential yield gains on the table and farmers in these areas less productive and poorer than they need be. We develop a simple theoretical model that illustrates how a confluence of demand and supply factors can result in such an underserviced equilibrium. To study the empirical veracity of this model, we study the disruption of the maize seed market in Western Kenya that took place when a combination of public sector foundation breeding and social impact investment capital allowed a local seed company to expand and target a niche agro-ecological zone with adaptively-bred maize varieties. A three-year randomized controlled trial reveals that the impacts of these seed varieties on farmer yields and revenues in the niche market were substantial, both for better-resourced farmers (who used nonadapted hybrids and fertilizer prior to the intervention) as well less well-resourced farmers (who did not). Taken together, this theoretical and empirical evidence suggests news ways for thinking about seeds systems in areas typified by the high levels of agro-ecological heterogeneity found in important parts of sub-Saharan Africa.