Aim To evaluate the surgical seminars given for medical students at Auckland University and factors affecting them, so as to define areas for improvement. Methods A confidential questionnaire was completed by fourth-year medical students for each surgical seminar they attended during an academic year. The seminars were repeated four times during the year. The questionnaire consisted of five statements relating to whether the seminar was stimulating, problem- oriented, patient-centred, evidence-based and interactive. The students rated each statement on a 7-point scale. Results 556 questionnaires evaluating 28 clinical teachers were collected. A generalized linear model showed that the clinical teachers (F=4·16, P = 0·0001), the attachment (F=4·39, P=0·005) and the interaction between the seniority and institute (F=6·38, P=0·019) had a significant effect on the rating of the seminars. University consultants and hospital registrars had the highest overall least-squares mean, followed by hospital consultants, while university lecturers had the lowest overall least-squares mean (5·56, 5·4, 5·18, and 4·36, respectively). Ratings for problem-oriented approach improved over the studied period, while ratings of evidence-based and patient-centred approaches remained lower. University consultants and hospital registrars had the best patient-centred approach compared with hospital consultants and university lecturers (least-squares means 5·52, 5·4, 5·01, 4·18, respectively). Hospital registrars showed the best interactive ability ratings (least-squares mean 5·51) while university lecturers had the least stimulating seminars (least-squares mean 4·46). Conclusions Surgical seminars which were introduced to cover surgical topics had acceptable ratings in each of the teaching domains, although there is a need for improvement in the patient-centred and evidence-based approaches. [ABSTRACT FROM AUTHOR]