Among the core concerns with the extraordinary proliferation of land deals in Africa - often referred to as "land grabs" - is that the signing of contracts between host states and foreign companies and/or other states for large swaths of territory and associated agribusinesses could represent an erosion of the host state's sovereignty powers. This concern reveals a double characterisation of the state, as weak in its sovereignty and, yet, as very able to negotiate and implement deals. Host states have been shown to be able to exercise sovereignty in those deals, what type of sovereignty - and whose -, however, remains in dispute.This thesis seeks to address this issue through a case study that focuses on the question how sovereignties are shaping and being shaped by land deals in Mozambique's Limpopo Valley. It specifically investigates the rice and sugar projects in areas of the Chokwe and Xai-Xai regadios. It considers land deals as a set of processes for international-domestic negotiation of goals and funding, followed by processes in the areas of decision-making, policy-making, and project implementation. Based on critical reappraisals of the concept of sovereignty, the thesis understands sovereignty as a set of powers that a state effectively has, beyond mere legal sovereignty, rather than an a priori attribute that a state does or does not possess, in zero-sum terms. As such it is an outcome of relational, inter-subjective processes and, thus, dynamic and historically contingent. Consequently, rather than absolute power over its territory and population, sovereignty is considered in terms of degrees of two types of political power practices, "command" power and "infrastructural" power, according to multiple and not always congruent state functions. To this, the thesis brings a notion of socially constructed state such that it is never neutral because a part of society and, thus socially embedded and produced. This allows me to move past the assumption of 'common good' and the moralist discussions of 'elite capture' and corruption.Based on this theoretical and analytical framework, the thesis posits irrigated agriculture and the state schemes hosting foreign projects as "sites" where actors' interests and powers are shaped relationally: the state (in different capacities), other states and their development agencies, foreign private sector actors and multiple domestic groups. The processes are studied at two levels. The first concerns how state "command" power is used to harness and/or defend against different international developments, negotiating international narratives and domestic needs, resulting in agricultural and water regulations, with ODA dependence for budgets. A subset of regulatory activity is the revisions to by-laws of management irrigation-scheme companies, as new representatives of central power locally. At the second level, the research focuses on interaction with Western equity and Chinese cooperation projects, two of the main types of investors, which come with different foreign management and funding models. Further, processes are embedded in historical trajectories of elite groups' moving away from agriculture since the 1980s, yet holding on to land entitlements, and of producers' displacement.This analytical framework allows research to effectively go beyond the notion of the state as either weak or able, considering it as polymorphous and acting in specific dimensions that no longer seem contradictory. Further, it illuminates the mutually constitutive nature of (sub)national and international dimensions of sovereignty, which tend to be exiled from each other in mainstream approaches to the notion, as well as the inextricability of political and economic powers in the 'sovereignty frontier' of post-conditionality states.