Abstract
Five consultant psychiatrists were trained in the use of two rating scales for assessing interview skills. The consultants rated 21 interviews recorded on videotape of psychiatric registrars assessing a 'patient' in a psychiatric setting. Inter-rater reliability was examined for all measures. Overall there was little agreement on most items when the raters were asked to give a qualitative judgement on a skill. Good agreement between raters was found when they were asked whether or not a question or action had been performed. In spite of the lack of agreement on many qualitative measures, four of the five raters did not differ significantly from each other in their global ratings.