This paper explores a potential relationship between rainfall data and household selfreported harvest shocks and local (spatial) variability of harvest shocks and coping strategies based on a survey of 2700 rural households in the Kagera region of northern Tanzania. In addition, correlations of rainfall amounts across the districts in the region are estimated in order to assess the variations in rainfall patterns across the districts. The results show that rainfall patterns in the region are very location-specific, there are few correlations between rainfall events, and that the distribution of household reported harvest shocks differs significantly between districts and correspond to the observed variability in local climate patterns. Coping strategies are focused on spreading risks and include reduced consumption, casual employment, new crops, external support and the selling of assets. There are no large differences in applied coping strategies across the region, but district-level data demonstrate how local strategies differ between localities within the districts. The results emphasize that in order to target rural policies and make them efficient, it is important to take into account the local conditions that rural households face when experiencing climate-related shocks. Finally, shocks reported by households appear to correspond well with observed variability in rainfall patterns.