This thesis is a theoretical and empirical investigation of the escalation process that results when states are in dispute over some issue(s) in the international system. Escalation is viewed as a cost-imposing bargaining strategy enacted by states for the purpose of eliciting concessions from the adversary. Each state has a cost tolerance which identifies the maximum costs a state is willing to suffer to achieve their demand on the issue at stake. A player determines the likelihood that escalation will produce the adversary's concession by comparing the cost tolerances of both players. Once a player decides that the opponent has a greater willingness to bear the costs of escalation, that player concedes and quits the escalation game. A formal theory of the escalation process is developed which produces several hypotheses identifying the conditions under which states are expected to escalate in conflict and, once the decision to escalate is made, the level of escalation that that state is likely to achieve. One advantage in developing a general theory of escalation is that the interrelationships among the variables are expected to hold across a wide variety of escalation contexts. These hypotheses are then tested empirically in order to assess the utility of the general theory of escalation. In one of these contexts, the extended deterrence crisis, potential attackers are expected to refrain from pursuing higher levels of escalation when the attacker can perceive the value that the defender attaches to its ally and when the defender has the capacity to impose serious escalation costs on its opponent. Another of the hypotheses tested produced the expectation that lower cost tolerant actors were more likely to achieve their highest level of escalation on their first move in the conflict given that they were unable to prevail in a long, drawn-out game of escalation with their stronger opponent. The results of the empirical tests indicate that the general theory of escalation developed in the thesis is useful in identifying the conditions that motivate states to escalate, the level of escalation that is likely to be achieved and the conditions under which states concede.