In this paper, I propose to reconsider some basic notions of Mamluk political organisation that underpin most research in Mamluk studies, but that remain too often still premised on awkward, unspoken or even unconscious assumptions about the concept, the nature and the institution of the late medieval Syro-Egyptian Mamluk polity. What exactly do we mean when we refer to 'the Mamluks' and 'Mamluk rulers', to 'the Mamluk state' and 'Mamluk authority', or indeed to 'the Mamluk empire' and 'Mamluk territory' ? How can we avoid vague, static or even anachronistic understandings of the complex late medieval realities for which these terms are being used ? Can we break free from related ideas of the historical exceptionalism of the Mamluk 'slave state', and reconnect with wider late medieval and early modern studies, where debates on state formation have been at the forefront of very rich academic debates? Can we simultaneously reconnect these debates with more genuinely Mamluk Arabic terms, such as kh???a, dawla or tadb?r ? These are some of the key questions that have informed this paper. In considering them, this paper problematises standard assumptions of the relationship between Mamluk elites and the 'state', and it offers one possible alternative model. Informed by ongoing discussions in related fields of history (Weber, Tilly, Reinhard, Bourdieu), by advances in Mamluk studies (Lapidus, Hodgson, Holt, Chamberlain, Clifford, …), and by Ibn Khaldun's historical model of a political organisation's transformation, this paper focuses in particular on the issue of Mamluk state formation, of how elites' monopolies of legitimate violence and resource appropriation were organised over time. In doing so, this paper will develop and present the idea of a 'Flux and Reflux'-model of Mamluk state formation, of Mamluk political organisation's continuous oscillation between and emanation from a set of paradigmatic types, derived from the the basic notion of the Mamluk Military Patronage State. With this new model in mind, this paper will then conclude by reconsidering the above mentioned notions of Mamluk elites, state and empire, and suggest how they may be thought in more meaningful Mamluk ways. In this way, this engagement with historical sociology offers some crucial reflections on the nature of the Mamluk polity, with a specific focus on the dynamics of the relationship between rulers and ruled. This will substantially add to stimulating the application of nuanced ideas about the perceptions and realities of any Mamluk empirial experience.