For over four decades seed system development in developing countries has followed a western (industrialized country) model which is based on market-economic thinking and in which maximization of crop productivity is the dominating goal. On the basis of this model, using estimated crop area and ideal seed replacement rates, farmer demand for seed is calculated. The perceived existing or projected farmers demand for quality seed thereby becomes the conditional factor in the development of a sustainable seed sector. Throughout time, the underlying rationale for these seed system development interventions has been unchanged: if only farmers would become good seed users , and regularly purchase seed, the seed sector will develop towards a more evolved state. While accounting for the constraints that smallholder farmers experience, demand building interventions still focus on convincing farmers of the benefit of and providing them with the possibilities to invest in good quality seed . We draw from a range of case studies on interventions in root, tuber and banana seed systems in Africa and others to point out the lack of understanding on what farmers demand for seed entails, and how this leads to unrealistic seed system interventions. Using case study evidence, we show how this focus leads to the promotion of services and technologies (e.g. micro credit, insurance, improved seeds and varieties) that do not fit the reality of the poorest farmers. We consider how promoting seed demand in this way may also promote particular interests, while also excluding certain groups of farmers. Implications for the concept of demand for seed and for seed system intervention will also be addressed.