Working in teams is widely recognised as an important skill across a range of occupations and professions, with the teaching of teamwork now a feature of the higher education curriculum (Drake et al, 2006). The success of teamwork, however, depends to a large extent on the mediated management of complex variables such as individual conduct, collective action, technologies, space and communication. Leathard and McLaren (2007) make the further point that the concept of teamwork may itself be subject to contradictory ideas about the meaning of teamwork and how teams ought to be structured. It is against this background that there has been an increase in the use of teams for doctoral supervision, in part reflecting a growing trend towards interdisciplinarity as part of the twenty-first century knowledge economy (Manathunga et al, 2006) and also as recognition that it is unlikely that a single supervisor will have the full range of knowledge and skills to support complex doctoral work. Specific methodological skill as well as subject knowledge is relevant and Phillips and Pugh (2000) discuss arrangements put in place by some universities whereby supervision involves external academics working jointly with internal ones. Thus supervision teams of two, three and four members are not uncommon and the charge to doctoral students made by Rugg and Petre (2004) to manage their supervisor(s) is therefore more challenging in light of the team dynamic that may not be cohesive or harmonious. This article considers some of the challenges and benefits of the practice of team supervision within postgraduate research education focusing on roles and relationships within the team and on the issue of conflicting views and expectations amongst supervisors that can both stem from and give rise to role tensions.