ECHA writes back: The reasons for rejection of read-across in compliance check decisions by the European Chemicals Agency.

Read-across is the most widely used adaptation to standard information requirements under REACH and is central to efforts to reduce unnecessary animal testing. However, its regulatory acceptance in European Chemicals Agency (ECHA) compliance check (CCH) decisions remains limited. This study systematically examined common deficiencies and trends in reasons that read-across adaptations fail based on publicly available CCH decisions. We reviewed ECHA CCH decisions published from 2010 through 2025, extracted information on read-across type, adaptation study type proposed, source and target substances, and reasons for rejection. The latter were coded using 17 assessment elements related to structural similarity, toxicokinetics, toxicodynamics, substance identity, and supporting data. Among 2,722 analyzed CCH decisions, nearly half contained at least one read-across adaptation, but only 85 read-across hypotheses were deemed acceptable. In contrast, 1,526 rejected read-across hypotheses were identified, most of which used analogue rather than group type of reasoning. Rejections were especially common for higher-tier repeated dose studies. The most frequent deficiencies in proposed adaptations involved unresolved uncertainty in toxicokinetic and toxicodynamic similarity, including limited evidence on bioavailability, metabolite formation, non-common compounds, and endpoint-relevant bridging data. Structural similarity was nearly always part of read-across but was rarely sufficient on its own. These findings show that successful read-across requires more than a plausible similarity argument; it must provide endpoint-specific evidence that residual uncertainty is low enough for regulatory use. Our results identify recurring pitfalls in submissions and support a path toward uncertainty-characterized, NAM-supported read-across that can improve regulatory confidence while reducing potentially unnecessary testing in animals.