Sir

Anonymous peer review is the fundamental process by which the scientific community evaluates grant proposals. One way in which the Internet has changed the review process is that granting agencies now allow, and some encourage, the construction of websites by authors to supplement proposals. Although these sites provide useful methodological and other information to reviewers, they threaten reviewer anonymity.

Server software can log the unique IP address of each computer that visits a particular website. These IP addresses can identify the institution and, in some cases, the name of the visitor. This is especially problematic when reviewers are the only visitors to a site created specifically for the proposal and hosted by the author, who has disclosed the URL only in the proposal. Reviewers who are aware of the risk of exposing their identity in this way may avoid websites associated with proposals, effectively rendering their review incomplete.

We contacted several funding agencies (NSF, USDA, NERC and the Wellcome Trust), and none has a policy regarding the use of websites in conjunction with grant proposals, nor do they have mechanisms to protect the anonymity of reviewers visiting proposal websites.

In the short term, we suggest that reviewers visit such sites using a dial-up Internet service provider (for example, America Online and NetZero) that will not provide any information about the reviewer's identity. As a long-term solution, granting agencies should consider hosting such websites rather than allowing submitting authors to host them at their home institution.

Whereas this Correspondence is primarily concerned with the grant review process, the same problem applies to articles submitted to journals for publication, if supplementary material essential to evaluate the article is supplied for peer-reviewers at an author's own website.