In the rural Tropics, the participatory risk assessment, based on local knowledge only, is very widespread. This practice is appropriate for hazard identification and for raising the awareness of local communities in relation to the importance of risk reduction, but it is still imprecise in determining risk level, ranking and treatment in a context of climate change, activities in which technical knowledge is unavoidable. Integration of local and technical-scientific knowledge within the framework of an encoded risk assessment method (ISO 31010), could favour more effective decision making with regard to risk reduction. The aim of this chapter is to verify the applicability of a multi-risk local assessment-MLA which combines local knowledge (participatory workshop, transect walk, hazard and resource mapping, disaster historical profile) and scientific knowledge (climate downscaling modelling, hazard probability and scenarios, potential damages, residual risk). The test is carried out in two villages of the Western Niger, particularly exposed to flooding and agricultural drought. The risk (hazard probability * potential damages) is identified, analysed (level of risk) and evaluated (residual risk, adaptation measures compared with potential damage costs). The MLA is feasible. The two villages, while bordering on one another, have a different risk ranking. Depending on the village, the risk treatment could reduce the risk level to 17 and to 41% of the current risk, with costs equating to 34 and 28% of the respective potential damages.