Why are some counterinsurgency campaigns effective while others are not? Existing theories give primacy to hearts-and-minds strategies, preponderance of state forces, and type of rebel group as determinants of effectiveness. This dissertation argues that variation in effectiveness results from differences in select tactical, organizational, and technological capabilities of the counterinsurgent state, captured by the concept of Legibility and Speed-of-Exploitation System (L&S;). The L&S; varies in the legibility of the population (legibility) and the speed of exploitation of legibility gains (speed). This dissertation shows that campaigns attaining high L&S; are effective even when failing to implement a hearts-and-minds strategy, fighting resilient insurgents, and without adequate force levels. Further, the dissertation shows that the variation in the L&S; is shaped by the domestic political salience of the counterinsurgency effort and prior experience of counterinsurgent forces in internal war. The dissertation tests predictions of the new theory and orthodox theories in multiple counterinsurgency campaigns and conflicts using novel fieldwork data, archival materials, and historiography: Indian counterinsurgency in Punjab (1984-1994) and Jammu and Kashmir (1989-1999), Pakistani counterinsurgency in the North West Frontier (2002-2011), British counterinsurgency in Malaya (1952-55) and Kenya (1953-56), US counterinsurgency in Iraq (2007-2010), and US Drone War in Pakistan (2004 to 2014). The dissertation’s theory and findings challenge the conventional wisdom that success in counterinsurgency is tied to winning the support of the civilian population.