This article uses two unique panel data sets to study the causal effect that armed conflict has over firm exit in Colombia. Using a fixed-effect estimation methodology at the plant level and controlling for the possible endogeneity of armed conflict through the use of instrumental variables, we find that a one-standard deviation ( SD) increase in the number of guerrilla and paramilitary attacks in a municipality increases the probability of plant exit in 5.5 percentage points or .28 SD. This effect is stronger for younger manufacturing plants, with a smaller number of workers and low levels of capital.