The arenaviruses are primarily viruses of rodents, but some members of the group cause severe disease (Argentine and Bolivian haemorrhagic fevers and Lassa fever) when transmitted to humans in the specific areas of the world where they are enzootic. Current research of relevance to the provision of vaccines against these diseases, which highlihts many of the problems encountered generally in the developpement of vaccines, is review here. Although one the classical approaches to vaccine production, the use of inactivated preparation of virus of varying degrees of purity has produced no results of promise, attenuation of a virulent strain of Junin virus by passage in cultured cells has yielded a vaccine strain currently being tested efficacy in protecting against Argentine haemorrhagic fever in the human population at risk. The experimental evidence for protection in animal model system by related, apparently non pathogenic, viruses and by recombinant vaccinia virus expressing arenavirus proteins is discussed, together with some of the potential difficulties of these approaches.