Résumé : This paper studies the impact of democratic transitions on institutional outcomes in a panel of 135 countries over the period 1984-2012, using an event study method. Our estimates suggest that the bulk of the improvement occurs during the three years following the transition. We can find no anticipation effect in average institutional outcomes. The results are robust to using alternative transition definitions and alternative codings of pre- and post-transition years, to changing the set of control variables, to excluding former socialist countries from the sample, and to dealing with endogeneity with IV regressions. When distinguishing full and partial democratic transitions, we find that both improve institutional outcomes. We find that the effect of democratic transitions is conditional on GDP per capita, education, and the regularity of the transition. When looking at specific components of institutional quality, we find that law and order, internal conflict, government stability, and ethnic tensions follow the general trend, while corruption, military in politics, bureaucratic quality, and investment profile indices are insensitive to democratic transitions.