BACKGROUND The emergence of medical students' professional identity is important. This paper considers this in a snapshot of the early years of undergraduate medical education. From the perspective of social identity theory, it also considers self-stereotyping, the extent to which individuals associate with attributes identified as typical of groups. METHODS Paper questionnaires were completed by first and second year medical students following teaching sessions at the beginning (October) and end (April) of the academic year. Questionnaires consisted of scales measuring the strength and importance of identity and self-stereotyping, referent to 'doctors' and 'students'. Linear mixed effects regression considered longitudinal and cross-sectional effects of progress through the course, and differences in responses to 'doctor' and 'student' measures. RESULTS In October, responses were received from 99% (n = 102) and 75% (n = 58) of first and second year cohorts respectively, and in April from 81% (n = 83) and 73% (n = 56). Response rates were over 95% of those present. Linear mixed effects regression found that all 'doctor'-referent measures were higher than 'student' measures. Strength of identity and self-stereotyping decreased between beginning and end of the year (across both groups). Men indicated lower importance of identity than women, also across both groups. There were no differences between year groups. Self-stereotyping was predicted more by importance of identification with a group than by strength of identification. CONCLUSIONS Findings reinforce observations that medical students identify strongly as doctors from early in their studies, and that this identification is greater than as students. Decreases over time are surprising, but may be explained by changing group salience towards the end of the academic year. The lack of a gender effect on strength of identification contrasts with the literature, but may reflect students' lack of 'performance' of professional identity, while the effect on importance is speculated to be linked to social identity complexity. Identification with professional group may have implications for how medical schools treat students. The findings on self-stereotyping have relevance to recruitment if applicant populations are limited to those already internalising a stereotype. There may be consequences for the wellbeing of those who feel they cannot fulfil stereotypes when in training.