While the heat wave impacts on public health have been widely addressed in developed countries especially after the intense event over West Europe during summer 2003, no effort has been made to detect them and evaluate their impacts in least developed countries, and especially Africa where the climate is warmer and adaptation capacities are low. Over West Africa preliminary interviews, climate and epidemiologic analyses show however that this problem is emerging. Moreover climate projections in this area indicate that such events should increase in frequency and intensity in the near future. However these climate models display huge biases in their mean state over this region, and recent studies highlighted large radiative and low-level temperature biases. In order to provide robust information on the future evolution of heat waves, it is necessary to reduce as much as possible these biases. Starting from this context, the main objective of ACASIS is to set-up a pre-operational heat wave warning system over West Africa tailored to health risks of the population living in this region. This is a demonstration project focused on Senegal and Burkina where national weather services have already started developing products dedicated to weather/climate and health relationships, and where several health and demographic observatories have been operating for up to several decades. Based on qualified meteorological, climate and demographic data bases, firstly, the dynamics of the heat wave events and their atmospheric patterns will be determined, as well as their evolution over the last decades. Their predictability at short and medium ranges will be evaluated on ensembles of multi-models forecasts outputs. On a longer time scale, control simulations and climate scenarios of the CMIP5/AR5 database will be analysed and the simulated future evolution and associated uncertainty of these events will be evaluated. More precisely the processes at the origin of model radiative biases will be examined and reduced as much as possible. In parallel, epidemiologic studies associated with interviews will be conducted in the health and demographic sites in Senegal and Burkina in order to evaluate the physiologic and social vulnerability of the African population to high temperature extremes. It will allow to define tailored bio-meteorological indicators to be used in the warning system. From these outcomes and by implementing downscaling to link the synoptic scale of the heat waves to local bio-meteorological indicators, we will set-up a demonstration warning system on a “testbed” platform named MISVA, already implemented as the result of a prior collaboration between Meteo-France, OMP and ANACIM, the meteorological agency of Senegal. Based on the interviews, and with the setting of several workshops with stakeholders and public institutions, we will be able to provide specific recommendations associated to these warnings. An implementation in the Meteo-France operational system at the end of the project or after might be possible. To carry on this project, a pluri-disciplinary consortium has been set-up gathering climatologists, physical processes specialists, meteorologists, biostatisticians, demographers, socio-economists, epidemiologists, geographers, and operational meteorological agencies. It will work through a close collaboration between French and African teams where young African researchers will be highly involved.