It is often assumed that transfers received from governments, nongovernment organizations (NGOs), friends and relatives help rural households to pool risk. In this paper I investigate two functions of transfers in Ethiopia: risk pooling and income redistribution. Unlike most of the literature this paper investigates not only whether but also how much risk pooling is achieved. I find evidence that transfers from governments/NGOs play a role in insuring covariant income shocks, (weak) evidence that transfers from friends/relatives insure idiosyncratic income shocks and evidence that transfers target the poor households. However, the contributions of transfers to risk pooling and income redistribution are economically very limited.