We argue that extreme positioning benefits parties in new democracies, because-given the lack of other reliable cues such as party histories-the distinctiveness of their left-right policy positions increases voter certainty about parties' identities and intentions in office. Cross-sectional analyses provide evidence that, in the new democracies of postcommunist Europe, parties that are farther away from the mean voter position gain more popular support than those moderately positioned along a policy continuum. In established democracies, by contrast, policy moderation increases popular support. We also find empirical support for the proposed causal mechanism that links policy positions to popular support via voter certainty. These findings have implications for party strategies, spatial theories, and our understanding of political representation in new democracies.E xisting cross-national research suggests that parties gain votes by adopting positions close to the mean (or median) voter's position (Alvarez, Nagler, and Bowler 2000a;Alvarez, Nagler, and Willette 2000b;Dow 2001Dow , 2011Schofield 2004; see also Schofield, Sened, and Nixon 1998a;Schofield et al. 1998b). 1 In this article, we argue that this relationship is less likely to hold in new democracies, where parties are, instead, more likely to be rewarded for extremism than centrism. Our argument, which we detail in the next section, builds on three central premises derived from the existing literature. First, parties with less ambiguous positions (i.e., positions about which voters, on average, are more certain) attract more electoral support. Second, centrist positions generate more ambiguity and voter uncertainty than noncentrist positions. Third, noncentrist or extremist positions are particularly effective in reducing voter uncertainty when other cues about party identities and intentions are absent, as is the case in new democracies.In order to test our argument, we conduct a macrolevel analysis of party positioning and election outcomes across 31 countries from 1996 to 2007. 2 Our study produces the following findings. First, with respect to the established democracies, we corroborate the findings of the existing studies of multiparty competition that parties receive a statistically significant electoral benefit from locating near the mean voter position. Second, in line with our theoretical expectation, we find that in the context of new, postcommunist democracies, parties receive an