Using panel household survey data from rural Ethiopia, we investigate informal risk sharing against health shocks in the presence of multiple risk sharing networks. We find that neither short-term nor long-term health shocks are insured through transfers from networks such as friends, neighbors, and members of informal associations. However, networks related along bloodline such as extended family members provide assistance when health shocks are long-term such as disabilities. The results show that these networks strategically complement planned component of their transfers which are made on a regular basis such as remittance, entitlement, or chop money (small cash sums for household expenses). Moreover, we find significant history dependence in transfers from not only genetically distant networks but also extended family members as well as formal institutions, which seems to discourage dependency. Finally, the findings suggest significant heterogeneity in transfers. 100 risk sharing against transitory illnesses versus long-term shocks, such as disabilities. The data comes from four rounds of Ethiopian Rural Household Survey (ERHS) panel data, covering about 1,480 households in 15 rural villages between 1994 and 1997. We consider transfers from different social networks, including family members, relatives, friends, neighbors, and members of informal savings, credit and funeral associations as well as formal religious, government and non-government organizations. Based on genetic proximity (along bloodline) to a household, we consider all possible networks which have made cash or in-kind transfers to the household.The dependent variable is cash or in-kind transfer, with large proportion of zeros which arise due to either a corner solution where individuals decide to make zero transfers or transfers are not in the choice set. Using Dynamic Correlated Random Effects Seemingly Unrelated Regression (DSUR) Probit and Tobit models, we address issues of non-linearity in the distribution of transfers, state dependence, unobserved individual heterogeneity, and initial conditions problem. Moreover, the DSUR models allow for transfers from one network to be correlated with the other, providing important evidence on the extent and direction of interactions between social networks. The approach captures the interactions among networks not only on the time-varying idiosyncratic components but also on the time-invariant components of transfer. While the former can be interpreted as unplanned transfers made in response to unforeseen events or idiosyncratic shocks, the latter can be interpreted as planned transfers made on a regular basis, such as remittances, entitlement, and chop money (small cash sums for household expenses). In implementation, we use the hierarchical Bayesian estimation method with Markov Chain Monte Carlo (MCMC) simulation and data augmentation techniques to estimate the models.To preview our results, we find that close family members and relatives, who are more likely to be altruistic along bloodline, make...