Why do some non-democratic regimes tolerate mass protest, while others do not? I investigate this question through a study of state response to protest in Azerbaijan and Belarus between 2002 and 2004. I argue that Azerbaijan's toleration and Belarus' limiting of public dissent reflect different strategies for maintaining political stability. This project develops a theory of protest toleration in non-democratic regimes, suggesting that authoritarian regimes can tolerate protest in order to gather information on citizens' true policy preferences. I then draw from literature on international and domestic influences on non-democratic rule to hypothesize several factors that shape the relative costs and benefits of tolerating versus limiting protest. These hypotheses will be tested using an original dataset of protest events and archival and interview research conducted in the field between August 2008 and May 2009.