This work explores empirical cases of a duopoly of violence in Latin America where Organized Criminal Groups (OCGs) and State governments negotiate to decrease high levels of violence. I theorize three different Models under which this negotiation in a duopoly of violence can be understood: a crackdown approach negotiation, a bribery negotiation, and a legitimate negotiation. By studying historical cases in El Salvador, Mexico, Colombia, Honduras, and the United States, I argue a legitimate negotiation led by States can be more effective and less economically and socially costly alternative to decrease high levels of violence. I then answer the question under what conditions do Latin American States negotiate with criminal groups? Historical case studies show levels of corruption, levels of violence, state capacity, and OCG capabilities are all correlated with one another. Results show large variation across variables and case studies. However, the general trend is that countries with medium to high enforcement capabilities are more likely to implement a crackdown approach negotiation, countries with high levels of corruption are more likely to engage in bribery negotiations, and countries with medium State enforcement capabilities or medium to high OCG credibility are more likely to engage in a legitimate negotiation. This thesis concludes with Latin American States' policy implications from negotiation, a discussion of governments' simultaneous usage of various negotiation models, and the need for further research for a duopoly of violence and governance.