Education on rational prescribing is receiving increased emphasis but the effectiveness of this education is not well researched. This study used a randomised trial to evaluate a seminar for general practice trainees on rational prescribing of antibiotics and benzodiazepines, two important areas of general practice prescribing. Results show a decrease in prescribing of antibiotics by the group of trainees attending the seminar but no effect on an already low level of benzodiazepine prescribing. The study provides evidence that group educational approaches to influencing prescribing behaviour can be effective.