A comparison is made between two styles of syndicate work. In the first, referred to as 'unstructured' (U), groups range in size from four to six students. In the second, referred to as 'formal' (F), groups are eight to ten in size and tend to adopt a formal chairman-secretary structure. This latter type is here illustrated by one of the authors from his experience in a Medical School, where the tasks presented to the students are complex 'problem boxes' requiring many man-hours from different departments for their creation. The analysis of the U type is based on the second author's experience in an American university and an English polytechnic. The areas examined are, first, the size and the composition of the groups; second the tutor's role; third, the development of higher order cognitive skills in the students; and fourth the management of research in this field. Questions are raised particularly in relation to the consolidation of syndicate work, to ensure command of the basic material; to the change of role required of both teachers and students; to the influence of the institutional culture on the work; to the adaptation of assessment techniques to measure development in the higher order skills; and to the possibility that investigation needs to be carried out on actual courses by teachers, students and researchers in collaboration.