Background: Covert duplicate publication is widely viewed as wasteful of peer reviewer and publishing resources and, if not detected and properly accounted for, can lead to biased estimates of effect sizes in systematic reviews. Objectives: To determine the number of covert duplicate publications in journals indexed in PubMed and uncovered within the systematic reviews of one Cochrane review group, the Skin Group, and to compare this list with those uncovered by Deja vu (http://discovery.swmed.edu/dejavu/), a new online tool for detecting duplicate publications indexed in PubMed. Methods: Covert duplicate publications were detected by electronically scanning the text of the 32 published systematic reviews of the Skin Group (http://www.csg.cochrane.org/about/pubdsysrevs.htm) for the word ‘duplicate’ and examining articles mentioned as duplicate publications. The Cochrane Skin Group Editor was also asked to identify duplicate publications (HC Williams). To match the contents of the Deja vu database, we defined covert duplicate publications narrowly as studies describing the same results without cross referencing them with abstracts in English in PubMed. Articles in journal supplements were excluded. Results: Two sets of covert duplicates were detected. Three of the four publications appeared in the Journal of the American Academy of Dermatology (JAAD), a high impact clinical journal with the highest h-index of any dermatology journal (http://www.scimagojr.com/journalrank.php?area=2700&category=2708&country=all&year=2006&order=h&min=0&min_type=cd). None of the four publications were initially identified by the Deja vu database. Restricting the Deja vu analysis to the 200 most likely duplicate publications in the JAAD (http://discovery.swmed.edu/dejavu/article/?q=J+AM+ACAD+DERMATOL) still failed to identify the three covert duplicate articles, but, interestingly, revealed two new sets of exact duplicate publications within the journal. Both sets of exact duplicates have been uniquely cited multiple times. The editor of the JAAD was contacted and admitted administrative error on the part of the journal. He has requested our assistance in writing a retraction of the exact duplicate articles that incurs no blame or penalty to the authors. Conclusions: The detection of covert duplicate publications by systematic reviews is complemented, but not replaced, by the new online tool Deja vu.