This article explains tactical escalation by a Peruvian left-wing group during the 1980s and 1990s as an interaction effect between organizational ideology and the broader political and organizational environment. In 1980, Peru's Sendero Luminoso (Shining Path) organization ended a decade of political organizing and launched armed struggle against a new civilian government. Peru had been governed since 1968 by military officers, but popular pressure, including strong left-wing protests, had forced the military to cede control. In responding to democratization with revolution rather than electoral participation, Sendero broke with the rest of Peru's Marxist left. In 1983, Sendero again escalated its tactics, initiating a campaign of violent intimidation against Peru's legal left. By 1996, according to data assembled for this study, the group had selectively assassinated some 300 prominent Peruvian leftists. For theorists of revolutions and social movements, Sendero's tactical trajectory poses two important puzzles. First, many revolutionary theorists believe that transitions from authoritarianism to elections decrease armed insurgency. Why, then, did Peru's democratization provoke Sendero's escalation? Second, Sendero might well have been expected to cooperate with other left-wing groups, rather than to attack them so brutally. Why did Sendero choose an alternative path? The group's anti-left measures are all the more puzzling given the opposition they provoked among potential allies at home and abroad. This article explains Sendero's choices by drawing on political opportunity theory, theories of organizational competition, and the concept of declining protest cycles. Democratization can promote greater levels of strife if small but violence-prone groups fear marginalization in electoral politics. A dense left-wing social movement sector, moreover, can stimulate internecine competitive fighting if only some of the movement's members accept the legitimacy of national elections.