Gabapentin: the legal and safety record

The 2004 plea: the origin of the psychiatric prescription

Gabapentin was approved in 1993 for one thing: add-on therapy for seizures. In May 2004, Warner-Lambert pleaded guilty to two violations of the Food, Drug and Cosmetic Act and paid more than $430 million — including a $240 million criminal fine — over what the government called a widespread, coordinated national campaign to promote the drug for unapproved uses: bipolar disorder, pain, migraine, attention deficit disorder and more [1]. Central to the case was the manipulation of the scientific literature — publishing what supported the campaign, sitting on what did not.

In the Justice Department’s words, the company “falsely promoted Neurontin as effective for treating bipolar disease, even when a scientific study demonstrated that a placebo worked as well or better than the drug” [1]. A decade later, the civil track added more: in 2014 Pfizer agreed to pay $325 million to settle the insurers’ class action and related suits, with final court approval that November, as its own annual report records [2].

How far the campaign reached: in 2001, national survey data put 83 percent of gabapentin use off-label, the highest share of any drug studied [5]. A later analysis found off-label use “grew steadily until settlement,” then fell along with on-label use, and concluded that settlements “have been dwarfed by the financial gains” from off-label marketing; its title is that the prosecution “did not deter” off-label use [6]. The other direction is in the same paper: the settlement itself was followed by a decline.

What this does and does not mean today

Hold both halves. The origin of gabapentin’s psychiatric use is court-documented marketing fraud — and your prescriber writing it today is doing nothing illegal and possibly nothing unreasonable. Off-label prescribing is legal, common, and sometimes right. What the record changes is the burden of the question: for a drug whose psychiatric prescription volume was built by a guilty-plea campaign rather than by approval trials, “what evidence supports this for my condition?” is not a hostile question — it is the only question. Gabapentin has no psychiatric approval and therefore no psychiatric approval trial — it sits in the “no number, here’s why” section of our trial-length page for exactly that reason.

Scale is why this matters: 46.9 million US prescriptions and 11.5 million patients in 2024 [4] — the highest-volume drug in our library, and the most-reported drug in our FAERS pulls. The volume, the reports, and the label are all on our gabapentin page, caveats first.

What is not in the record

• No Class I recall (within openFDA's ~2012-onward coverage window). Twenty-five recall records, all Class II or III, for manufacturing and quality problems [3]. FDA’s most serious category has never been applied.
• No adjudicated injury verdicts on psychiatric use. The 2004 case was about marketing conduct and payer losses — not a finding that gabapentin harmed any particular patient. A page that lets you infer otherwise is overreaching, in either direction.

Common questions

What was the Neurontin settlement about?

In May 2004, Warner-Lambert — by then a Pfizer subsidiary — pleaded guilty to two violations of the Food, Drug and Cosmetic Act and paid more than $430 million over the off-label promotion of Neurontin (gabapentin). The government described a coordinated national campaign promoting the drug for bipolar disorder, migraine, ADD and other unapproved uses; gabapentin was approved only as add-on therapy for seizures.

Is it legal for my doctor to prescribe gabapentin for anxiety or mood?

Yes. Off-label prescribing is legal and common, and it is the marketing of off-label uses by manufacturers that was illegal. The practical question for you is not legality but evidence: what trials support gabapentin for your specific condition? For many psychiatric uses the honest answer is thinner than the prescription volume suggests.

Did Pfizer pay more after 2004?

Yes. Pfizer’s own 2014 annual report states that in April 2014 it agreed to settle the third-party payer class action and related individual payer suits over Neurontin marketing for an aggregate of $325 million, and that the federal court gave final approval in November 2014.

Has gabapentin had serious recalls?

No Class I recall — FDA’s most serious category — appears in the openFDA enforcement database (openFDA enforcement data is reliable only from roughly 2012 onward, so this is a statement about the covered window, not the drug’s whole history). Twenty-five records exist, all Class II or III, for manufacturing and quality issues rather than anything about the molecule.

Does gabapentin have a black box warning?

No. Gabapentin has never carried a boxed warning — FDA’s most prominent label warning — and our medication page renders no boxed-warning section for it because none exists. What it does have: a 2019 FDA warning about serious breathing difficulties when gabapentinoids are combined with opioids or other CNS depressants, or used by patients with respiratory risk factors. A warning in the label and a boxed warning are different instruments, and pages that blur them are rounding up.

Is gabapentin dangerous?

That is a different question from anything on this page, and the record here does not answer it. What is documented: a 2019 FDA warning about respiratory depression when gabapentinoids are combined with opioids or in patients with respiratory risk, and growing scheduling at the state level. Bring the combination question to your prescriber if it applies to you.

Sources

Last verified 2026-09-29. Corrections change this date.