Since much of the focus on large-scale land acquisitions is predominantly political and ideological, different models and practices embedded in the phenomenon and, by consequence the diverse implications they imply, tend to be overlooked. This is supported by the use of the term "land grabbing": while it implies large differences in forms of organization of the production, investment processes and outcomes these land deals might take, the existing body of literature misses the economic and institutional nuances of investment models embedded in "land grabbing". The objective of this paper is to present the diversity of investment models implemented in Southern Africa and to analyze their differentiated implications in the framework of the region's broader agrarian trajectories. Based on intensive empirical research in Southern Africa and using a theoretical framework based on institutional economics (focusing on the institutionalized forms of agricultural production, the investment implementation processes and the extent of the implications), the results show that beyond the classical institutionalized forms of agricultural production (independent commercial farming, estate farming) new investment and production models are developing in the region. Six models with several sub-models have been identified: independent farmers (independent farmers, delocalized auxiliary farm model, Resource pooling farmers), cooperative, 1,000-day speculative, asset management, contracting (nucleus-estate, reverse tenancy, ingrower schemes) and agribusiness models. Besides the lack of inclusive business models, another important commonality of these models is the high failure of the investments - unless strongly integrated structures and value-chains are developed - leading to the establishment of few corporate structures. The paper reflects on Southern Africa's agrarian transformations, which, although not broad-based, are mainly characterized by the imposition of a dominant corporate-based paradigm. (Résumé d'auteur)