Households in rural Kenya are sensitive to weather shocks through their reliance on rain-fed agriculture and livestock. Yet the extent of vulnerability is poorly understood, particularly in reference to extreme weather. This paper uses temporally and spatially disaggregated weather data and three waves of household panel survey data to understand the impact of weather extremes –including periods of high and low rainfall, heat, and wind– on household welfare. Particular attention is paid to heterogeneous effects across agro-ecological regions. We find that all types of extreme weather affect household well-being, although effects sometimes differ for income and calorie estimates. Periods of drought are the most consistently negative weather shock across various regions. An examination of the channels through which weather affects welfare reveals that drought conditions reduce income from both on- and off-farm sources, though households compensate for diminished on-farm production with food purchases. The paper further explores the household and community characteristics that mitigate the adverse effects of drought. In particular, access to credit and a more diverse income base seem to render a household more resilient.