After independence, cases of poisoning -- or at least of accusations of poisoning -- have undergone an impressive recrudescence in South Malekula, as elsewhere in Vanuatu and in many post-colonial situations. Some commentators link the phenomenon to a context of political insecurity and instability in which authority and its forms are questioned (Rodman 1993, Geschiere 1988). Others have a more generic approach and see in sorcery the necessary process of identifying the intentionality of social perturbations (Clément 2003). In most of these studies, be it from a historical and political or from a cognitive perspective, sorcery is seen as a correlative of unstable and disruptive situations, or as producing such situations; if not both at the same time. The aim of this paper is to question these positions. Sorcery may also involve strategies and processes of redistribution and of reorganisation of political legitimacy and practice in local contexts. Thus, while accusations of poisoning may be disruptive for the individual, this is not systematically the case from a sociological point of view. Following Bloch's study of sorcery among the Zafimaniry of Madagaskar (1998), I argue that sorcery or poisoning in South Malekula are directly tied to questions related to the domestic unit and its change over time and in space; and in particular the change of its area of political and discursive influence. What is more relevant in anthropological terms than sorcery and associated practices of poisoning themselves, are the public announcements and the identification of accusations of poisoning. In a context in which legitimacy is questioned and where traditional and imposed forms of leadership collide and are redefined, accusations of poisoning seem to be deployed when processes of coalition and competition reorganise the frontiers of domestic groups and their positioning in the political landscape.