Warlords have proven remarkably resilient, in Afghanistan and elsewhere. According to Malejacq the key lies in their ability to build multiple forms of power (ideological, economic, military, social, political), which they project effectively but sparingly, and to convert one form of power to another when needed. Their ability to operate at multiple levels, from the local to the international, is also critical. Building a Weberian state through external intervention is bound to fail, as warlords flexibly maintain their power by providing identification, security, income and various services in return for loyalty. The book offers a comprehensive discussion of warlords and develops a typology of their interaction with state structures (which they at times are part of), emphasizing 'what they do (their actual behavior) instead of what is assumed of them' (p. 4). Three rich case studies of Afghan warlords – Ismael Khan, Rashid Dostum and Ahmed Shah Massoud, as well as his successor Fahim Khan – nicely illustrate differences in their power portfolios and individual strategies of adaptation. In the 2001 US-led intervention, warlords provided the bulk of the boots-on-the-ground. A few years later, when foreign statebuilders wanted to minimize their political influence, the warlords had dug in. Despite the marginalization of the warlords under the Taliban's de facto rule (1996–2001), Malejacq takes their return for granted and advises external interveners to accept their critical role in governance and incorporate them, as confrontation proves counterproductive. This is an implication he draws for all situations where warlords stand strong. Among the many studies of warlordism and the failure of Afghan statebuilding, Malejacq's book stands out – conceptually succinct, empirically rich, and well crafted. It deserves a wide readership.