American civil discourse suffers from the incivility of its rhetoric and the relative disengagement, ignorance, and bias of its citizenry. Without the space or motivation to discuss serious issues in a sober tone, discussion of serious topics devolves into name calling, sloganeering, and a general avoidance of the problems facing the country. Deliberative democratic theory - a normative model of democracy in which citizens engage in effortful, unbiased, reason-based deliberations with one another in service of finding and implementing a common good - has been advanced as a possible remedy to our civic shortcomings (Gutmann & Thompson, 2004). Extant research has shown that properly constructed deliberative environments increase participants' topic-specific knowledge and alter participants' attitudes. The study detailed here extends this line of research by examining deliberation's ability to induce complex attitudes, those with both a positive and negative evaluation of the attitude object. Further, it tests deliberation's ability to do so relative to non-deliberative alternatives in both a general sample and among those likely most and least ready to engage in deliberation. A process model of complexity induction and maintenance is presented and deliberation's relative ability to retain the complexity induced is assessed. Results indicate that deliberation yields comparable or lesser degrees of intra-attitudinal complexity in the short term and no advantage or deficit in the long term. Implications for the study of deliberation and measurement of intra-attitudinal complexity are discussed.