Learning language in the classroom often means interacting with peers in small groups. There is a growing body of empirical research highlighting the factors that have an impact on student interaction and an increasing focus on the social context. This article aims to broaden the focus of research by considering the influence of emergent leadership on group interaction and opportunities for language learning. Japanese university students in an oral English course were placed into random small groups, fixed for a semester. The researcher collected video observation data of student classroom interaction twice during the semester, during midterm and final group conversation tests. Students rated the leadership of group members, and interactions were analyzed by selecting individuals representing strong leadership, moderate leadership, and weak or no leadership. The researcher conducted a regression analysis to determine the relative influence of English proficiency, extraversion, and leadership on behavioral engagement. Language-related episodes also were examined to determine cognitive engagement. Results show that leadership rather than proficiency or extraversion predicted participation in conversation, and cognitive engagement seemed to be negatively associated with strong leadership. The implications of emergent leadership for researchers and teachers are discussed.