This study extends existing research on political discussion’s influence on political knowledge in two ways. First, it expands the measures of discussion-related cognition to include discussion elaboration and perspective taking. Second, it employs panel data, which permit stronger causal inferences than cross-sectional studies. Our findings indicate that, even controlling for prior knowledge, interest, news use, and news elaboration, political discussion frequency and discussion elaboration are positively related to political knowledge. However, perspective taking is unrelated to political knowledge.