Introduction: Cochrane reviews are intended to be used by clinicians, consumers and policy makers as a source of evidence of effects of interventions. However, the methods by which priorities for reviews are determined are not explicit and do not appear to involve end-users of reviews. As Cochrane Review Groups depend on the time and commitment of usually unpaid reviewers this is a powerful force in allowing priorities for reviews to be determined by reviewers themselves.

Discussion: The Cochrane Collaboration is now at a point in its development where its ability to operate internationally and produce high quality reviews is not in doubt. It needs now to take stock of needs and priorities of relevant stakeholders. The Canadian Cochrane Centre, in collaboration with the Canadian Heart & Stroke Association, leading Canadian academics and the Cochrane Heart Group held a meeting in Ottawa, 19 October 1998, to discuss priority setting. Three inter-related groups were identified that might play a role in priority setting, as follows:

* Funders and consumers asked whether it was logical for review priorities to be set only by the academic community to which most reviewers belong.
* Clinicians, consumers and policy makers are unlikely to make much use of material - however good - in which they have made no investment.
* As funding is generally sought from government, charities, and end-users of the Cochrane Library, it seems essential that these "stakeholders" should have a say in the priorities to be set.

This model represents a potentially valuable means of defining priorities. The Cochrane Heart Group is intending to develop this approach, initially within the UK as the editorial base is located here. If stakeholder meetings prove to be useful, they will be extended to cover other geographic regions where there is interest in working more collaboratively to define review priorities.