The purpose of this teaching experiment was to examine the causal effect of relevant (compared to irrelevant) instructor self-disclosure on student affect and cognitive learning. Undergraduate students (N = 288) were randomly assigned to a 19-minute classroom lecture with an instructor who taught the same lesson but self-disclosed either relevant or irrelevant information while teaching. Results indicated that relevant instructor self-disclosure increased student affect in students' likelihood to enroll with the instructor again but did not influence students' general affect toward the instructor. Findings also revealed a direct effect of self-disclosure relevance on students' test scores; on average, students scored 8.42% higher on a short-term recall test compared to students in the irrelevant self-disclosure condition, controlling for lesson coherence. Practical implications are provided for how instructors might use relevant self-disclosures in their teaching.