The effect of specific teaching of interviewing techniques, measured by a rating scale, was compared in experimental and control groups of trainee general practitioners. The subjects were given the opportunity to observe their own consultations on audio-visual recordings. No significant overall effect was demonstrated, but more detailed analysis revealed trends that suggested that the effect of the teaching only occurred in the latter half of the training year and was accompanied by an increase in the time of the consultation. It is our conclusion that the study should be repeated with larger numbers. It is possible that the absence of any effect of teaching interviewing techniques in the early part of the training period could be due to the traumatic experience of moving from hospital practice to primary care.