Sub-Saharan Africa (SSA) remains the world's most food-insecure region characterized by high levels of child mortality and poverty and low levels of human and physical capital (FAO, 2009). Countries in SSA, including Tanzania, heavily depend on a smallholder-based agricultural sector, which makes their welfare and food security particularly vulnerable to climate change (Barrios et al., 2008). The goal of this study is to provide a comprehensive analysis of the impact of weather risk on rural households' welfare in Tanzania using nationally representative household panel data together with a set of novel weather variation indicators based on interpolated gridded and re-analysis weather data that capture the peculiar features of short term and long term variations in rainfall and temperature. In particular, we estimate the impact of weather shocks on a rich set of welfare indicators (including total income, total expenditure, food expenditure and its share in total expenditure and calorie intake) and investigate whether and how they vary by different definitions of shocks - capturing changes in levels and variations over different time periods. We find that both rainfall and maximum temperature variability exert a negative impact on welfare (i.e. no consumption smoothing) and that households that have adopted SLM practices are able to achieve income-smoothing. We also find that the most vulnerable rural households are much more affected by a rainfall deficit compared to the households in the top income quantile. Results underline the key role extension services play in enhancing adaptive capacity to reduce vulnerability to adverse weather conditions, as well as the importance of targeting the most vulnerable households in policy interventions to improve food security in the face of weather shocks.