Index-based insurance has emerged as a compelling strategy for agricultural risk management in Africa, particularly in contexts where smallholder farmers are disproportionately exposed to climate-related hazards. Despite its potential, the effectiveness of this mechanism is often constrained by the spatial heterogeneity of crop yield responses to climatic variables – an issue that significantly contributes to basis risk. This study investigates the value added by spatial econometric approaches, with a focus on the geographically weighted panel regression (GWPR) model, in enhancing the calibration of index-based insurance schemes under conditions of spatial instability in the relationship between maize yields and climate indicators in Burkina Faso.