A new Primer in PLOS Biology warns that poorly validated antibodies can silently mislead biomedical research, with a large-scale test finding more than half of evaluated antibodies unfit for at least one intended application. Johan Duchêne of Ludwig-Maximilians-University Munich and coauthors describe how antibodies can bind off-target proteins while still producing “biologically plausible” staining patterns. The paper cites real examples where earlier findings were contradicted by independent challenges, contributing to wasted samples, time, and animal and human tissue use. It emphasizes that validation is application-specific, since performance in Western blot assays may not translate to immunohistochemistry or other readouts. The Primer points to initiatives such as YCharOS (Antibody Characterization through Open Science), which systematically tests commercial antibodies using knockout samples as a gold-standard approach for specificity.