• education, measurement/*standards;
  • edu-cation, medical undergraduate/*standards;
  • *learning;
  • peer review;
  • curriculum;
  • reproducibility of results;
  • Brazil

Aims  To examine the features of cross-year peer tutoring and to explore their relationships to learners' characteristics and educational outcomes from the student-tutor perspective.

Method  The records of 447 final year medical students were examined to provide data on the starting terms, frequency and course targets of peer tutoring activity of student tutors. The relationships of these features with their learners' characteristics, academic achievements and selective clerkship pathways were analysed.

Setting  The medical education programme at the University of Brasilia, Brazil.

Results  Analysis showed that about 96% of all graduates had acted as student tutors at some time during the programme, with great variation in starting terms, numbers and types of courses tutored. The average number of tutored courses per tutor was four. Frequency and variety of tutored courses were significantly related to achievement, learning style and gender. Higher achievers acted as student tutors for many terms and explored different subjects, and there is evidence that the experience expanded their academic expertise. Specific tutoring in a clinical course also related to strength of early career preference. Furthermore, there was a significant correlation between the number of terms of tutoring undertaken in a clinical course and the proportion of students choosing selective clerkship training in the same area by the end of programme.

Conclusions  The findings suggest that acting as a peer tutor can be an appealing and constructive educational opportunity to further students' academic development. Enhanced expertise seems to relate to the accumulation and breadth of tutoring experience. Moreover, clinical tutoring may help students in making decisions regarding choice of career.