Small-group discussion is a central component of 21st-century biology classrooms. Many factors shape these discussions and thus influence potential learning gains. This study examined how accuracy and idea consideration shaped small-group discussions in undergraduate biology labs (12 groups, M = 42.8 talk turns). To do this, we asked: (1) Is there a relationship between a student's science accuracy and the amount peers consider the student's ideas? (2) To what extent does peer consideration of a student's ideas predict that student's ability to steer the conversation? Building on this second question, we then explored (3) Does general group academic ability or immediate conversational accuracy better predict group learning? To answer these questions, we coded aspects of discourse (science accuracy, idea consideration, etc.) before quantitative analysis. Strong correlation was found between students' science accuracy and idea consideration (r = 0.70). Both accuracy and idea building predicted one's ability to steer the conversation. Subsequent analysis highlighted the critical role of immediate discourse in group learning. Group-level analysis revealed that group performance was not related to the group's overall ability in the classroom, but rather the immediate accuracy of their group conversations. Implications and limitations are discussed.