Democracies rarely if ever fight one another, but they participate in wars as frequently as autocracies. They tend to win the wars in which they participate. Democracies frequently build large alliances in wartime, but not only with other democracies. From time to time democracies intervene militarily in ongoing conflicts. The democratic peace may contribute to a normative justification for such interventions, for the purpose of promoting democracy and eventually for the promotion of peace. This is reinforced by an emerging norm of humanitarian intervention. Democracies may have a motivation to intervene in non-democracies, even in the absence of ongoing conflict, for the purpose of regime change. The recent Iraq War may be interpreted in this perspective. A strong version of this type of foreign policy may be interpreted as a democratic crusade. The paper examines the normative and theoretical foundations of democratic interventionism. An empirical investigation of interventions in the period 1960-96 indicates that democracies intervene quite frequently, but rarely against other democracies. In the short term, democratic intervention appears to be successfully promoting democratization, but the target states tend to end up among the unstable semi-democracies. The most widely publicized recent interventions are targeted on poor or resource-dependent countries in non-democratic neighborhoods. Previous research has found these characteristics to reduce the prospects for stable democracy. Thus, forced democratization is unpredictable with regard to achieving long-term democracy and potentially harmful with regard to securing peace. But short-term military successes may stimulate more interventions until the negative consequences become more visible.