In this study natural-in-action metacognitive activity during the student laboratory in university physics is explored, with an aim towards quantifying the amount of metacognition used by the students. The study investigates whether quantifying natural-in-action metacognition is possible and valuable for examining teaching and learning in these contexts. Video recordings of student groups working during three types of introductory physics laboratories were transcribed and then coded using a coding scheme developed from related research on mathematical problem solving. This scheme identifies a group's general behaviour and metacognitive activity. The study recognizes that reliably identifying metacognition is challenging, and steps are taken to improve reliability. Results suggest that a greater amount of metacognition does not appear to improve students' success in the laboratory--what appears to matter is whether the metacognition causes students to change behaviour. This study indicates that it is important to consider the outcome of metacognition, not just the amount.