Explaining learning contents to a fictitious peer (i.e., non-interactive teaching) improves learning, yet this effect is modest, heterogeneous, and likely influenced by individual differences. We examined whether the effectiveness of non-interactive teaching could be increased by incorporating drawing or distributing teaching. We realized a 3 × 2 field experimental design (N = 317), crossing the factors learning activity (restudy, teaching-only, teaching + drawing) and timing (after the study phase or distributed three times throughout the study phase). Overall, teaching resulted in better immediate conceptual knowledge than restudying, mediated by the level of completeness. This teaching effect was most pronounced in the after-study condition. However, drawing did not enhance conceptual knowledge. Students who taught underestimated their immediate knowledge. No lasting effects were observed. Students with higher academic self-concept or work ethic benefited more from teaching, highlighting the moderating role of inter-individual differences for instructional interventions. Educational relevance and implications statement: This classroom study demonstrates that non-interactive teaching is an effective instructional method in secondary school physics education. The findings highlight the importance of considering students' individual differences, such as academic self-concept or work ethic, when designing such learning activities. These insights emphasize the need for adapted and differentiated approaches that can better account for individual differences, ensuring that non-interactive teaching can be effective across diverse student populations.