Patent enablement is the requirement that your patent document teach a person skilled in the field how to make and use the full scope of what you claim, without undue experimentation. It comes from 35 U.S.C. § 112(a), and it is the price of the patent bargain: the public grants you exclusivity for twenty years, and you hand the public a real working teaching in return. Claim more than you taught, and the claim falls, even when the invention itself indisputably works.
The same statutory sentence carries a second, separate demand. The specification must also contain a written description of the invention showing you actually possessed it when you filed. Enablement asks whether you taught it. Written description asks whether you had it. Courts treat those as different questions, and a claim can satisfy one and fail the other.
The statute requires the specification to contain “a written description of the invention, and of the manner and process of making and using it, in such full, clear, concise, and exact terms as to enable any person skilled in the art to which it pertains” to make and use it. It also requires the inventor to set out the best mode contemplated for carrying out the invention, though the America Invents Act of 2011 removed failure to disclose the best mode as a ground for invalidating or holding a patent unenforceable. Best mode survives as a filing duty with almost no enforcement teeth.
Three obligations, then, packed into one sentence: written description, enablement, best mode. The first two do essentially all the litigation work.
The measure of enablement is not the invention. It is the claim. If you claim a genus, you must enable the genus. If you claim only the species you built, you are safe but narrow. This is the central tension in patent drafting: attorneys are paid to claim broadly, and breadth is exactly what makes a claim vulnerable under § 112(a).
Amgen Inc. v. Sanofi, 598 U.S. 594 (2023), is the modern statement of the rule and the case worth understanding first. Amgen had isolated antibodies that bind a protein called PCSK9 and stop it from degrading the LDL receptors that extract LDL cholesterol from the bloodstream. It described 26 such antibodies by their amino-acid sequences and gave the three-dimensional structures of two. Then it claimed the whole class: every antibody that binds the specified region and blocks the interaction, however structured.
What the specification offered for the rest of that class was a roadmap: make antibodies, screen them, keep the ones that work. A unanimous Court, in an opinion by Justice Gorsuch, held that this was not enablement. Reaching back to nineteenth-century authority including O’Reilly v. Morse and the incandescent lamp cases, the Court reasoned that if a patent claims an entire class, the specification must enable the entire class. “A specification may call for a reasonable amount of experimentation,” the Court wrote, but Amgen’s disclosure amounted to instructions to run the same trial-and-error search the inventors had run. The claims were invalid.
The decision did not create a new test. It reaffirmed a very old one, and it did so in the field where broad functional claiming had become routine.
Some experimentation is always permitted. A patent is written for a skilled person, not a novice, and that person is expected to fill gaps using ordinary knowledge in the field. The line is between routine work and open-ended research.
The Federal Circuit’s framework comes from In re Wands, 858 F.2d 731 (Fed. Cir. 1988), which lists eight factors:
The factors are illustrative, not a checklist, and no one of them decides a case. Two of them do most of the pulling. Predictability explains why a mechanical claim of enormous breadth survives while a chemical or biological genus claim of similar breadth does not: in an unpredictable art, one working example predicts little about the next compound. Breadth explains why the same disclosure can enable a narrow claim and fail a wide one. Notably, Wands itself came out for the applicant. Screening hybridomas was routine in that field, so the experimentation was not undue.
For years, litigants argued that written description was just enablement stated differently. The Federal Circuit ended the argument, sitting en banc, in Ariad Pharmaceuticals v. Eli Lilly, 598 F.3d 1336 (Fed. Cir. 2010). Section 112, the court held, contains two distinct requirements, and the test for written description is whether the disclosure reasonably conveys to a skilled artisan that the inventor had possession of the claimed subject matter as of the filing date.
Ariad’s claims covered methods of reducing NF-κB activity in cells, a functional result. The specification proposed classes of molecules that might achieve it but described no actual working molecule. The court invalidated the claims: describing a desired outcome and hypothesizing how to reach it is not describing an invention. As the opinion put it, a patent is not a hunting license.
The practical bite of the possession requirement shows up in prosecution rather than at trial. When an applicant amends claims during examination to cover a competitor’s product, the new claim only gets the original filing date if the original specification described that subject matter. Written description is what polices that. It is also the doctrine behind rejections of claims added to capture something the inventor learned about years after filing.
Three consequences follow, and they are in tension with each other:
Breadth is bought, not asserted. A genus claim is only as good as the disclosure supporting it. In unpredictable arts, that means working examples across the range, structure-function relationships that let a skilled person predict what else falls in the class, or a structural feature common to the genus rather than a functional one.
Functional claiming is the exposed flank. Defining an invention by what it does rather than what it is invites the Amgen problem directly. It is not forbidden, and § 112(f) expressly permits means-plus-function claiming, but a claim to “any compound that achieves X” needs disclosure that reaches every compound achieving X.
Everything is fixed at filing. You cannot add new matter to a pending application. Whatever teaching the specification lacks on the day it is filed, it lacks permanently, and a continuation inherits the same defect. This is why the disclosure decision, and not the claim decision, is the one that determines how much protection is available later.
What is patent enablement? Enablement is the requirement in 35 U.S.C. § 112(a) that a patent’s written description teach a person skilled in the field how to make and use the claimed invention without undue experimentation. The teaching has to reach the full scope of the claim, not just one working example. A claim that covers more than the specification enables is invalid, even if the invention itself works.
What is the difference between enablement and written description? Enablement asks whether the specification teaches a skilled person how to make and use the invention. Written description asks a different question: whether the specification shows the inventor actually possessed the invention as of the filing date. The Federal Circuit held in Ariad v. Eli Lilly (2010) that these are two separate requirements in the same statutory sentence, so a claim can be fully enabled and still fail written description.
What are the In re Wands factors? Eight considerations the Federal Circuit set out in In re Wands (1988) for judging whether experimentation is undue: the quantity of experimentation needed, the amount of direction or guidance given, the presence of working examples, the nature of the invention, the state of the prior art, the skill of those in the art, the predictability of the art, and the breadth of the claims. No single factor controls; they are weighed together.
What did Amgen v. Sanofi decide about enablement? In Amgen Inc. v. Sanofi, 598 U.S. 594 (2023), a unanimous Supreme Court invalidated antibody claims defined by function rather than structure. Amgen described 26 antibodies by amino-acid sequence but claimed the entire genus that binds a target and blocks it. The Court held that disclosing a trial-and-error screening process is not enabling the class, because the specification must let a skilled person reach the full scope without undue experimentation.
Going further: What Is Patentable, and what is not .
This page is general legal information, not legal advice, and it does not create an attorney-client relationship.
The Federal Circuit set out the eight-factor test for whether practicing a claim demands undue experimentation, the USPTO's enablement standard today.
In 1895 the Supreme Court voided Sawyer and Man's claim to all fibrous incandescing conductors. The case remains the ancestor of full-scope enablement.
A unanimous Supreme Court invalidated Amgen's antibody patents for failing to enable the full scope of what they claimed. The decision revives a demanding, century-old conception of the patent bargain with particular force in the life sciences.
The en banc Federal Circuit confirmed that Section 112 demands a written description separate from enablement, invalidating Ariad's NF-kB patent for claiming a result it never showed it possessed.
Reversing a $1.2 billion judgment, the Federal Circuit held that a CAR-T patent claiming any antibody binding element failed written description because it disclosed no representative species or common structure for the vast scFv genus.
The Supreme Court sustained Samuel Morse's telegraph patent but voided his eighth claim to every use of electromagnetism for printing at a distance as too broad and unsupported.
A specification that disclosed a single rapamycin compound could not support claims reaching tens of thousands of structurally diverse molecules, the Federal Circuit held, because finding the active ones required excessive screening.
The Federal Circuit erased the largest patent verdict in U.S. history, holding that a functional nucleoside genus spanning billions of candidate molecules was neither enabled nor adequately described.
