This paper explores risk sharing in the Zone Lacustre, Mali, as viewed through the lens of consumption smoothing. We find that idiosyncratic shocks appear to have little impact on consumption, and that households respond to these shocks in a variety of ways. In general, nonpoor households are more likely to enter into new income-generating activities while poor households are more likely to engage in credit or gift exchange or to ration consumption. When we construct a stronger test for consumption smoothing, we find that changes in household income lead to modest changes in consumption. Covariant shocks, as measured by village/round dummies, always lead to changes in consumption. These results are robust to concerns regarding bias resulting from measurement error or endogeneity of changes in income. Lastly, we find that households with access to improved water control infrastructure are less vulnerable than those that rely on rainfall or the flooding of the Niger River.