The Retraction Watch Database — the field’s most comprehensive public record of retracted scholarly papers, now maintained by Crossref following its 2023 acquisition of the project — has passed 63,000 logged retractions. Retraction Watch’s own year-in-review post, published December 30, 2025, states plainly: “The Retraction Watch Database now includes 63,000 retractions.” That figure is a running total accumulated since the database’s earliest records (retractions dating back to the 1970s and 1980s have been backfilled where documentation exists), not an annual count — but it marks a milestone worth noting for anyone who tracks the pace of correction in scholarly publishing.
What the 63,000 figure does and doesn’t tell you
A rising cumulative total is, on its own, an ambiguous signal. It can reflect more misconduct and error entering the literature, or it can reflect better detection and a stronger institutional willingness to act on it — most researchers who study retraction trends, including Retraction Watch’s own founders, argue the honest answer is some mix of both. The database’s growth has been shaped by a handful of large, discrete events as much as by a steady background rate: Wiley’s 2023-2024 retraction of more than 11,000 papers tied to its Hindawi acquisition, for instance, is visible as a sharp spike in Retraction Watch’s own decade-by-decade retraction charts, not a smooth trend. Anyone citing the database’s size as evidence of a change in the underlying rate of research misconduct should look at rate data — retractions relative to total publication volume — rather than the raw cumulative count alone.
A real number, from a real hearing: what Congress was told about the retraction rate
In April 2026, Kate Travis, managing editor of Retraction Watch, testified before the House Committee on Science, Space, and Technology’s Investigations and Oversight Subcommittee at a hearing examining publish-or-perish incentives and scientific publishing practices. As reported by Inside Higher Ed, Travis told the subcommittee that “the overall retraction rate is about 0.2 percent” — framing that figure as evidence that science’s self-correction mechanisms are functioning, rather than as proof of a system in crisis. Her broader testimony, also covered on Retraction Watch’s own site, pressed lawmakers on a different problem: that researchers and outside investigators often cannot get access to the underlying data and materials needed to substantiate a misconduct concern in the first place, and she argued for reforming incentive structures that reward volume over careful, correctable work.
CASRAI is not able to independently verify a specific historical comparison figure (for example, a rate reported for 2016) that has circulated alongside the 0.2% figure in some secondary summaries of this story. We have not been able to trace that comparison to the hearing record, a Retraction Watch primary source, or another citable original — so this page reports the verified current-rate figure from the testimony as covered by Inside Higher Ed, and does not repeat the unverified historical comparison. If a primary transcript surfaces a sourced 2016 (or other historical) figure, this page will be updated.
Why the distinction between count and rate matters for research administrators
For anyone using the Retraction Watch Database operationally — screening a candidate hire’s publication record, checking a grant applicant’s track record, or vetting a manuscript’s reference list — the practical takeaway from both figures together is the same: retraction remains a rare outcome relative to total published output, but the absolute number of records to check against has grown substantially, and a single retraction can still carry serious consequences for the individuals and institutions involved. A low base rate does not mean the mechanism doesn’t matter to the specific cases that hit it.
Using the database
CASRAI’s guide, How to Search the Retraction Watch Database, walks through the practical mechanics of querying the database by author, journal, institution, or reason for retraction. For the editorial process a retraction actually goes through once a concern is raised, see How Long Does a Retraction Take? and CASRAI’s dictionary entry on Retraction Watch. Editors and authors handling the writing side of a retraction can also see CASRAI’s guide, How to Write a Retraction Statement. For a recent example of how the database is being used in original research, see CASRAI’s coverage of a cross-publisher analysis of retraction and reinstatement rates, Cross-Publisher Study: Elsevier Retracts Least, Reinstates Most Among Major Publishers.
