Intense rainfalls in Sub-Saharan Africa are increasing in frequency. Land degradation, watercourses siltation, and flood defence failure turn these events into disastrous floods. Over the last decade flood risk assessments have been prepared to face these disasters. However, they have frequent limitations in design, accuracy, and completeness. The objectives of this study are (i) to integrate local and scientific knowledge into a participated pluvial flood risk assessment (ii) to identify assets and (iii) to estimate the potential impact and efficiency of risk-reduction measures. The assessment is developed in four rapidly expanding towns of Niger, flooded several times in recent years. Flood-prone areas and assets are identified according four flood scenarios using local knowledge, 2D hydraulic modelling, and visual photointerpretation of very-high-resolution satellite images. Risk-reduction measures are singled-out through public participation. The residual risk and benefit/cost analyses provide a decision-making tool to accept or treat risk. During the last decade the expansion of the four towns has been more rapid in flood-prone zones than in safe areas. Nowadays more than half of the housing stock could be flooded by rainfalls with 20 years return period. Catchment treatment and building retrofitting can reduce risk. from 100 to 29–82. Nevertheless, the benefit/cost of risk reduction is high for towns settled in small catchments only.