We investigated the relation between providing and receiving audio peer feedback with a deep approach to learning within online education. Online students were asked to complete peer feedback assignments. Data through a questionnaire with 108 respondents and 14 interviews were used to measure to what extent deep learning was perceived and why. Results support the view that both providing and receiving audio peer feedback indeed promote deep learning. As a consequence of the peer feedback method, the following student mechanisms were triggered: “feeling personally committed,” “probing back and forth,” and “understanding one's own learning process.” Particularly important for both providing and receiving feedback is feeling personally committed. Results also show that mechanisms were a stronger predictor for deep learning when providing than when receiving. Given the context in which instructors face an increasing number of students and a high workload, students may be supported by online audio peer feedback as a method to choose a deep approach to learning. Lay Description What is already known about this topic: Higher education aims to promote deep approaches to learning among students and increasingly provides education online. From face‐to‐face education, we know that it is oral interaction that plays a major role in promoting deep approaches to learning and students perceive that as more personal than written feedback. But interaction in online education usually takes place in a written way due to the demand for student participation that is independent of space and time. The audio feedback in earlier studies had been provided by instructors, not by peers. Earlier studies also focused more on receiving peer feedback rather than providing feedback. What this paper adds: Both providing and receiving audio peer feedback lead to perceived deep approaches to learning. The following student mechanisms were triggered: “feeling personally committed,” “probing back and forth,” and “understanding one own's learning process.” Particularly important, both for providing and receiving feedback, is feeling personally committed. Results also show that the student mechanisms were a stronger predictor for a deep approach to learning when providing than when receiving feedback. We suggest that audio peer feedback makes great demands on feeling personally committed and as a consequence both feedback providers and feedback receivers choose a deep approach to learning. Implications for practice and/or policy: Results may help instructors identify where, how, and why audio peer feedback practices might be used to promote a deep approach to learning in online education. Instructors should provide sufficient instruction about how to provide feedback aimed for a deep approach to learning, and more specifically on how to record and publish and listen to audio peer feedback.