Editors' Summary: The OMB mandate for peer review of "information products" across the federal government, the Klamath Falls, Oregon, saga, and the legislative attempts to bind the regulatory arms of the ESA through peer review have sparked vigorous debate about the use of peer review in regulatory settings. In this Article, the first empirical treatment of the subject, J.B. Ruhl and James Salzman seek to show that regulatory peer review can meaningfully improve agency decisions that rely on the use or interpretation of scientific information, but that this alone does not determine whether peer review should therefore become part of agency decision processes. They caution that regulatory peer review may well prove unwise if there is no clear understanding of the real extent of the problem peer review is supposed to address. The authors propose an approach of randomized peer review to shift the debate away from whether regulatory peer review is good or bad, or whether agencies are biased or not, and on to a more productive, empirically grounded vantage from which we can more intelligently assess the proper role for this process in agency settings.