Peer tutoring is a potentially valuable resource in higher education. There are few published accounts of the impact of peer tutoring in medical education. College-wide experience of peer tutoring together with difficulties recruiting medical teachers for a communication programme led to the development of a peer-tutoring project. This paper reports the impact on first-year medical students of peer tutoring in patient-centred interviewing. After attending a preparatory workshop, third-year medical students co-facilitated their first-year colleagues in a session:Interviewing a Simulated Patient. First-year students completed written evaluations immediately after the session and two months later randomly selected students were assessed in patient-centred interviews. The impact of the peer-tutoring experience was evaluated by comparing these outcomes for students in groups co-facilitated by peer tutors with those who worked in groups facilitated by medical teachers. The eight learning objectives were completely met by more than 56% of students. However, there were statistically significant differences for four objectives with more students in groups facilitated by medical teachers completely meeting these objectives. Although the seven educational techniques used in the session were rated favourably by all students, two were rated as more effective in achieving the learning objectives by students in groups facilitated by medical teachers. Free-text comments revealed no differences between groups. Two months after the session, there were no differences between students in terms of interviewing skills as rated by trained observers and simulated patients, whilst simulated patients were more satisfied with interviews from students facilitated by peer tutors (p <0.02). Peer tutors can support the acquisition of basic patient-centred interviewing skills in first-year medical students when contributing to one session of a structured programme. First-year students were receptive and willing to learn from their senior colleagues.