Can terrorist attacks be timed to change the outcome of democratic elections? In this paper, we analyze the electoral impact of the terrorist attacks of March 11, 2004, in Madrid. Studies using individual level postelectoral survey data reach contradictory conclusions. We propose an alternative approach. Since the bombings took place only three days before the 2004 congressional election, we can find a control group of individuals who cast their vote before the terrorist attacks. The results indicate that the attacks had an important electoral impact, rejecting the hypothesis that the identity of the winner was unaffected by the terrorist attacks.