The Second Circuit Just Put a $40 Billion Product-Liability Claim Back on the Table—and the Science Was Already In

The Second Circuit is helping Kenvue’s buyers take $40 billion of someone else’s money.

The ruling itself is procedural: the court found that Judge Denise Cote, according to public court records the presiding judge over the multidistrict litigation, applied the wrong standard when she excluded the plaintiffs’ expert witnesses. Cote had ruled that Dr. Andrea Baccarelli, the dean of Harvard’s public health school, and the other proffered experts cherry-picked results and were unreliable. The Second Circuit said she was too aggressive—that the Daubert gatekeeping standard requires a trial, not a summary judgment on the experts’ credentials. The case goes back to the Southern District of New York for a fight over whether the experts were actually unreliable or merely contestable.

That procedural holding is perfectly defensible as a matter of federal evidence law. Daubert v. Merrell Dow Pharmaceuticals, 509 U.S. 579 (1993), and Kumho Tire Co. v. Carmichael, 526 U.S. 137 (1999), instruct trial judges to screen expert testimony for relevance and reliability—but the standard is supposed to be a gate, not a wall. The Second Circuit’s opinion tracks the language of Amorgianos v. National Railroad Passenger Corp., 303 F.3d 256 (2d Cir. 2002), which warned that Daubert does not authorize a trial judge to “rule for the defendant as a matter of law simply because the judge finds the plaintiff’s expert testimony unpersuasive.” Cote’s exclusion of Baccarelli and his co-experts may have crossed that line.

But here is the rub: the scientific record that the Second Circuit did not review—because it was not before it—is essentially closed. A Lancet review of 43 studies published in January 2026 found no causal link between acetaminophen use during pregnancy and neurodevelopmental disorders. That review was the largest meta-analysis on the question to date, pooling data from nearly a quarter-million mother-child pairs across three continents, according to the review’s authors. Its conclusion: “Current evidence does not support causal effects of acetaminophen use during pregnancy on children’s neurodevelopmental outcomes.”

The medical consensus anchored in that review is the status quo ante. The American College of Obstetricians and Gynecologists, the FDA, and the major medical societies have all continued to recommend acetaminophen as the safest over-the-counter pain and fever medication for pregnant women. The plaintiffs’ theory of the case requires a jury to believe that decades of institutional recommendations were wrong, that the Lancet review missed something, and that Baccarelli’s cherry-picked subset of studies identifies a signal the meta-analysis did not find.

That is not an impossible argument for a plaintiffs’ lawyer to make. Mass-tort litigation has a well-documented history of generating verdicts that the later scientific record does not support—the silicone breast-implant litigation, the talc-and-ovarian-cancer litigation, the birth-defect litigation against Zofran. Several of those cases produced plaintiffs’ verdicts or blockbuster settlements that the subsequent epidemiological record struggled to sustain. The plaintiffs’ bar is not in the business of waiting for the science to settle.

The difference here is the size of the target. Kenvue, the Tylenol manufacturer, is being acquired by Kimberly-Clark for more than $40 billion. The acquisition price was set before the Second Circuit revived 500-plus lawsuits. The probability that those lawsuits produce a judgment that materially alters Kenvue’s valuation is low—the science does not support the claim, and Cote’s original exclusion, while too aggressive for the Second Circuit’s Daubert standard, may well be vindicated on remand after a proper evidentiary hearing. But the cost of defending 500 lawsuits, even weak ones, is real. And the settlement calculus changes when the acquirer’s balance sheet, not the acquired company’s, is on the line.

The Second Circuit did what appellate courts do: it corrected a procedural error. It did not endorse the plaintiffs’ scientific theory. It did not hold that Baccarelli’s testimony was admissible—only that Cote needed to apply a different analytical framework before deciding whether it was admissible. The opinion is narrow, technical, and entirely consistent with the Daubert line.

But the effect is not narrow. Five hundred lawsuits that were dead are now alive. The plaintiffs’ bar has a Second Circuit opinion it can cite in every other acetaminophen case in the country. The settlement pressure on Kenvue—and by extension on Kimberly-Clark—just increased. And the acquirer, which priced the deal under one legal regime, now operates under another.

The science was already in before this ruling. The Lancet review was the definitive statement. The Second Circuit did not disturb it, because appellate courts do not review meta-analyses. They review trial judges. That distinction is real in law and invisible in practice. The practical effect is that a $40 billion transaction now carries a litigation tail that the best available evidence says has no merit, but that the federal judiciary has decided is a question for a jury, not for science.