Background: Collaborative teaching strategies are associated with enhanced student performance in quizzes when conveying complex physiological phenomena, (Giuliodori et al, 2006), and Mazur’s peer instruction may enhance transfer and retention of learning in a time-efficient fashion. This may be of particular relevance when conveying complex physiological phenomena, such as acid-base physiology. In the present study, we hypothesized that peer instruction was superior to conventional group work for teaching medical students to diagnose acid-base disorders by means of arterial blood gas analysis. Methods: A total of 41 second-year medical students were included during a respiratory physiology course. They were randomised to (A) conventional group work (n=22) or (B) peer instruction (n=19), and were tested by an identical pre- and post-test, which encompassed four arterial blood gases (see Table). Each correctly diagnosed acid-base disorder was awarded with one point (both tests). During the class, students were presented with eight additional arterial blood gasses that were either solved in groups of two (A; 2 minutes for each case) or by using a peer instruction-based approach (B; individual assessment for 1 minute followed by an assessment in groups of two for 1 minute). The percentage of correct answers and the subjective confidence level (0-100 % certainty of the correct diagnosis) in the pre- and post-tests were determined, and data are reported as median with corresponding interquartile range (IQR). Results: Pre-test scores were identical in both groups (median, 2; IQR, 1-2), and a significant increase was observed in both the conventional and the peer instruction study group following intervention (median, 4; 3-4 and 3; 3-4, respectively, both p < 0.0001), and scores did not differ between the two groups (p=0.31). Confidence levels increased in both groups and they evaluated the class similarly (data not shown). Conclusion: Collaborative teaching strategies were efficient for teaching students to diagnose acid-base disorders by means of arterial blood gas analysis. In the present time-restricted setup, we found no evidence that peer instruction is superior to conventional group work. However, a larger study is required before any definitive conclusion on this matter can be surmised.