We investigate the effect of coach dismissals on team performance in the Colombian Professional Soccer League. We use 5,600 observations encompassing first-division games played between 2003 and 2010. Because firing a coach might depend on team performance, we use the number of remaining matches in the season and the stage of the tournament as sources of exogenous variation in the probability of a coach dismissal. We find that firing a coach does not have a statistically significant effect on team performance measured by the number of points, the goal difference, the number of goals scored or the number of goals allowed. We relate our findings to the existing literature.