Nonviolent resistance can be a powerful tool for ordinary civilians to transform their governments; however, not all nonviolent movements end in success and many ultimately escalate into violent conflicts. To understand this escalatory process, I begin with the premise that social movements are not unitary actors but a collection of groups with varied preferences on goals and tactics. I argue that escalation is likely when movements have violence-wielding groups among their varied factions, as these groups deal in violence, believe in its utility, and can make the strategic decision to engage in violence as needed. I argue this is particularly likely when the campaign fails to make progress using nonviolent channels, suggesting that nonviolent tactics will not be successful to achieve the group’s goals. Expectations are tested using the Nonviolent and Violent Campaigns and Outcomes data and the case of Algeria’s escalation from a nonviolent movement to brutal civil war, and results are generally supportive.