The paper reviews the following three types of mangrove-shrimp systems: (i) integrated with canals between platforms planted with mangrove; (ii) associated having larger areas of water and a large mangrove area; and (iii) separated, with a dyke separating ponds from forest. The variations in shrimp yield of integrated and associated systems, that is, mixed systems, are attributable to water exchange, % water surface, primary production, stocking of post-larvae, leaf litter fall and decomposition, species, cover and age of mangrove, and predators. Leaf litter from all mangrove species except Nipa palm adversely affect water quality. Leaf composition and decomposition rate vary between species; submersed leaves decompose faster. Low concentrations of decomposing leaves of certain species temporarily boosted shrimp growth. Shrimp yield has been found to be highest in ponds with 30?50% mangrove cover, but remained long-term computation. An integrated mangrove-shrimp farm of 4 ha may provide farmers with a decent livelihood in Vietnam, if contracts do not restrict recommended practices to keep water quality within the limits acceptable for shrimp. Separated systems, where mangroves are fully connected to open water, have more potential to contribute to ecological, economic and social sustainability. Enhancing the economic returns by capitalizing on ecosystem services through carbon credits and organic certification, will require support for initial investment and clustering of farms.