The democratic peace theory, a long-accepted axiom of international relations, posits that democracies rarely, if ever, fight other democracies. This is attributed to a combination of shared values, institutional constraints, economic interdependence, and public accountability of leaders. Consequently, it is assumed that other regime types are less reliable in observing international peace. However, this generalisation fails to account for the many instances of non-bellicose autocracies. Hence, in recent decades the study of international relations has expanded to explore the possibility of a phenomenon called dictatorial/autocratic peace. The paper analyses literature that studies the conflict proneness of autocratic regimes toward other autocracies and democracies. It finds that, similarly to democratic leaders, authoritarians are faced with a variety of constraints in initiating or continuing wars. These constraints can be institutional, economic, and normative. There is evidence to support that peace is more likely between two authoritarian countries of similar regime types, especially dyads of single-party states. The research shows that autocracies face costs and limitations that should deter them from initiating war with democracies. It challenges the general assumption that democratic leaders are more accountable than autocratic leaders because they may lose the next elections. The paper finds that the domestic liability in case of engaging in a losing war for authoritarian regimes can sometimes be higher than in democracies, as autocrats face threats to their power and life.