Current research often emphasizes the typological convergence of violent non-state actors, which according to this literature tend to mix the characteristics of actors conventionally understood as criminals, insurgents, terrorists, or warlords. In this article, we examine the adequacy of such proposals to explain the evolution of violent actors in Latin America. Through a comparative study of the long-term development of two important Latin American violent actors, the Colombian FARC and the Mexican Sinaloa Cartel, we arrive at the conclusion that convergence phenomena occur with much less frequency and intensity than suggested, and when they do occur, they are caused by mechanisms typically not considered in the literature. This may indicate a significant limitation regarding the empirical reach of the convergence hypothesis as well as a need for further conceptual clarification.