Increasing levels of democratic freedoms should, in theory, improve women’s access to political positions. Yet studies demonstrate that democracy does little to improve women’s legislative representation. To resolve this paradox, we investigate how variations in the democratization process—including pre-transition legacies, historical experiences with elections, the global context of transition, and post-transition democratic freedoms and quotas—affect women’s representation in developing nations. We find that democratization’s effect is curvilinear. Women in non-democratic regimes often have high levels of legislative representation but little real political power. When democratization occurs, women’s representation initially drops, but with increasing democratic freedoms and additional elections, it increases again. The historical context of transition further moderates these effects. Prior to 1995, women’s representation increased most rapidly in countries transitioning from civil strife—but only when accompanied by gender quotas. After 1995 and the Beijing Conference on Women, the effectiveness of quotas becomes more universal, with the exception of post-communist countries. In these nations, quotas continue to do little to improve women’s representation. Our results, based on pooled time series analysis from 1975 to 2009, demonstrate that it is not democracy—as measured by a nation’s level of democratic freedoms at a particular moment in time—but rather the democratization process that matters for women’s legislative representation.