Cropping system is an essential aspect to take into account to manage blast disease (caused by the fungus Magnaporthe oryzae). In addition to the selection of resistant cultivars, studies report opportunities to limit blast incidence by managing mineral amendment (N, Si, P, etc.), cultivar mixtures or other cropping system adaptations. In Madagascar, rice is the staple crop and food. Farmers traditionally grow irrigated or rainfed lowland rice wherever possible. In the mid-1980s, CIRAD and FOFIFA launched a research program for the highlands to extend upland rice growing areas in high elevation areas of the tropics. This program was consolidated with research on cropping practices that ensure the sustainability of upland rice based cropping systems in this poor and fragile environment. New varieties where obtained, adapted for rainfed cropping up to 1800 m altitude (Dzido et al., 2004). However, farmers had to face attacks of blast disease. Due to the small genetic basis of these varieties, the fungus quickly overcame resistant or tolerant lines selected by breeders. The ways of management of blast disease are very limited in a country like Madagascar where efficient solutions are often not suitable for farmers. Observations of blast epidemics in different regions made us consider the soil as a key factor for rice susceptibility. In a first experiment, we measured the potential of our very susceptible varieties to tolerate blast when cropped in different soil conditions. During two years, we transported volcanic soil from a very fertile area close to the experimentation site and where blast pressure is much lower. We observed a decrease of blast symptoms on rice cropped on that soil compared to the soil of the experimental site, both on leaves and panicles. That decrease was bound to a significant yield increase, demonstrating the importance of plant nutrition on blast incidence. Such results confirm those obtained in 1981 by Seguy et al. Our aim was then to develop