Abstract: A common problem in conducting a systematic review is that appropriate data are not available from reports of clinical trials. A reviewer is then faced with the prospect of contacting the primary investigators to request such information. Exactly how one should go about this is not clear. Is it necessary to excite their interest first, or is it safe to ask for data at the outset? How should they be contacted? Is it worth putting great efforts into tracking investigators down? We present the protocol for First Contact, a randomised trial of two contrasting, but practical, approaches to contacting primary investigators. An intensive approach based on methods that appear to work based on evidence from randomised trials within population surveys will be compared with a simple request. We appeal to Cochrane reviewers uncertain of how to make first contact with trialists to join the trial. Central randomisation will take place at the UK Cochrane Centre and current results from a Bayesian analysis of the accumulating results will be available online at http:// www.mrc-bsu.cam.ac.uk/firstcontact to help reviewers both within and outside First Contact.