The laboratory notebooks of physics undergraduates taking two second-year practical courses were audited to discover whether they had used feedback comments in their subsequent coursework. Ninety-five per cent of the 37 students on the first course and 100% of the 14 students on the second course whose work was audited had used feedback. The marker's comments were classified into two groups based on whether they addressed simple (mastery) or complex (developmental) learning outcomes. Mastery comments were more likely to be acted on than developmental comments which aimed to extend students' skills and understanding to higher levels. This has implications for the use of feedback audit as a quality control process, since the feedback which is most commonly applied by students is not the most valuable for the development of higher order skills. Following reflection on the results for the first course, students taking the second course were given responsibility for checking their peers' notebooks against preset criteria. Peer checking improved students' marks but did not eliminate the need for mastery feedback. It is argued that a direct audit of students' use of feedback is particularly valuable when undertaken by the teacher who provides the feedback.