The war between the Colombian state and Pablo Escobar's Medellín cartel is perhaps the most well-known example of cartel–state conflict on record; this chapter marshals the theory developed in Part I to explain its dynamics. I make the following claims: (1) Colombia's initial crackdown and escalation (1983–1986) were unconditional. This led cartels—united under Medellín's leadership—to adopt increasingly violent strategies toward the state. (2) From 1987 onward, repression increasingly focused on Medellín to the exclusion of Cali because the former engaged in anti-state violence; that is, conditionality across cartels increased. This contributed to Cali's decision to attack Medellín (1988), and permitted it to operate in concert if not cooperation with the state to destroy Escobar (1992–1993). (3) The Sometimiento policy (1990–1991) allowed the state to credibly promise reduced repression on Medellín if it eschewed violence going forward, raising conditionality within cartels. Medellín members responded by eschewing anti-state violence. (4) The rescinding of Sometimiento after Escobar's escape (1992–1993) decreased conditionality within cartels, leading Escobar to re-engage in anti-state violence. I also conjecture that after the fall of Medellín and Cali, an effectively conditional approach has deterred cartel–state conflict. I describe the policy reform process that produced Sometimiento and the failed attempts at reform that preceded it, and discuss the unique role of violent lobbying in Colombia.