Table of Contents
You draft a claim that reads squarely on a competitor’s product. Then a court construes one phrase under 35 U.S.C. § 112(f), ties it to a single embodiment buried in your specification, and your broad-looking claim shrinks to almost nothing. That is the quiet danger of means plus function claims — limitations written as a function to be performed rather than a structure that performs it. Used deliberately, functional claiming is a legitimate tool. Used by accident, it is one of the fastest ways to lose scope or have a claim held invalid for indefiniteness.
What Are Means Plus Function Claims?

A means-plus-function limitation claims an element by what it does instead of what it is. The classic form is “means for [performing a function].” Congress expressly permits this style in 35 U.S.C. § 112(f) (formerly § 112, sixth paragraph), which says a claim element may be expressed as a means or step for performing a function without reciting the supporting structure.
There is a catch built into the same statute. When you claim in functional terms, the claim is construed to cover only the “corresponding structure, material, or acts described in the specification and equivalents thereof.” In other words, means plus function claims are not as broad as they read. Their scope is anchored to whatever structure your patent document actually discloses to carry out that function — plus a narrow band of equivalents.
So the phrase does double duty: it lets you claim a function, but it silently imports your specification’s embodiments as the boundary of protection. Understanding that trade-off is the whole game.
When Does 35 U.S.C. Section 112(f) Apply? The Three-Prong Test
For years, drafters assumed that avoiding the magic word “means” kept a limitation out of § 112(f). That assumption died in Williamson v. Citrix Online, LLC, 792 F.3d 1339 (Fed. Cir. 2015) (en banc), where the Federal Circuit scrapped the “strong presumption” that language lacking “means” is not a functional limitation. The test now asks whether the words “are understood by persons of ordinary skill in the art to have a sufficiently definite meaning as the name for structure.”
The USPTO applies a three-prong test in MPEP § 2181. A limitation invokes § 112(f) only when all three are met:
- Placeholder. It uses “means,” “step,” or a generic non-structural placeholder standing in for structure.
- Function. That placeholder is modified by functional language, usually linked by “for” or “configured to.”
- No sufficient structure. The claim does not itself recite adequate structure, material, or acts to perform the function.
Two presumptions ride on the word “means.” Use it, and the limitation is presumed functional unless the claim supplies its own structure. Omit it, and the limitation is presumed not functional — but that presumption falls apart if you used a nonce word plus a function with no structure. The label matters far less than the substance.
Nonce Words That Secretly Invoke Section 112(f)

The dangerous limitations are the ones that look structural but are not. Courts and the USPTO treat a set of “nonce” words as generic substitutes for “means” when they are paired with a function and no real structure. Common offenders include:
- mechanism for…, module for…, device for…
- unit for…, component for…, element for…
- member for…, apparatus for…, system for…
There is no fixed list. The MPEP is explicit that no placeholder always triggers § 112(f); context controls. Terms like “processor,” “circuit,” or “controller” can escape the statute when the specification and the art give them understood structure, yet the same word can fall inside it when the claim only recites a black box tied to a function. If you would not be comfortable pointing to a corresponding structure in your specification, assume a court will read the limitation as functional.
The Corresponding-Structure Trap for Software Claims
Once § 112(f) applies, your specification must disclose the structure that performs the claimed function, or the claim is indefinite under § 112(b). For mechanical inventions that is usually manageable — a spring, a lever, a gear train. For software it is a minefield.
The Federal Circuit has repeatedly held that a general-purpose computer is not enough. Under Aristocrat Technologies v. International Game Technology, 521 F.3d 1328 (Fed. Cir. 2008), and WMS Gaming v. International Game Technology, 184 F.3d 1339 (Fed. Cir. 1999), the corresponding structure for a computer-implemented function is a special-purpose computer programmed to perform a disclosed algorithm. If your specification does not spell out that algorithm — in prose, a flowchart, math, or steps — the limitation has no structure and the claim can be invalidated. A bare statement that “known techniques” can perform the function does not cure the gap.
This is why so many software patents drafted with “module for” or “engine configured to” language collapse in litigation. The claim invoked § 112(f), the specification never disclosed an algorithm, and the claim was held indefinite. A strong patent specification is the only backstop.
How Means Plus Function Claims Narrow Your Scope
Even when a functional claim survives, it is often narrower than the drafter expected. Because scope is limited to the disclosed structure plus equivalents, an accused product that performs the same function with a materially different structure may not infringe at all. That is the opposite of what most inventors want from a claim that, on its face, reads broadly.
There is also a subtle interaction with the doctrine of equivalents. The “equivalents” captured by § 112(f) are judged at the time of issuance and focus on structural equivalence to the disclosed embodiments, a distinct and often stricter inquiry than after-arising equivalents. The practical lesson: functional claiming trades breadth for a specific, structure-bound scope. Sometimes that trade is worth it. Often it is not.
Consider a simple example. Claim a “fastening means for securing two panels” and disclose only a threaded bolt, and a competitor who joins the same panels with an adhesive weld may fall outside your claim entirely — the function matches, but the structure does not. Had you disclosed bolts, rivets, clips, and adhesives, your equivalents range would be far wider. This is why disclosure strategy, not clever claim wording, ultimately decides how much a functional claim is worth. A prior-art picture that shows how competitors actually build the invention helps you disclose the right structures in the first place.
7 Smart Rules to Draft Means Plus Function Claims Safely

Whether you want to invoke § 112(f) or avoid it, do it on purpose. These rules keep functional limitations from becoming accidental invalidity grounds:
- Decide intent per limitation. Choose consciously whether each element should be functional or structural — never let a nonce word make the choice for you.
- Disclose real structure. For every functional element you intend, describe corresponding structure in the specification, and for software, at least one concrete algorithm.
- Link function to structure. Make clear in the specification which structure performs which claimed function; courts will not guess.
- Provide multiple embodiments. More disclosed structures means broader equivalents and a wider claim scope.
- Audit your nonce words. Search claims for module, unit, mechanism, engine, and device, then confirm each has structural support.
- Add structural claims in parallel. Pair any functional claim with independent claims that recite structure directly, so the portfolio does not rise or fall on § 112(f).
- Pressure-test before filing. Ask whether a litigator could point to the corresponding structure in five seconds. If not, fix the specification.
Handled this way, functional claiming becomes a scalpel instead of a landmine. For the broader discipline, see our guides to writing strong patent claims and claim drafting strategy.
How PerspireIP Can Help
Functional claiming rewards precision and punishes guesswork. PerspireIP’s patent professionals help inventors and counsel draft claims, build the corresponding-structure support your specification needs, and stress-test portfolios against § 112(f) before an examiner or a defendant does it for you. Contact us to strengthen your next application.
Frequently Asked Questions
Are means plus function claims still allowed?
Yes. 35 U.S.C. § 112(f) expressly permits claiming an element by the function it performs. The scope is just limited to the corresponding structure disclosed in the specification and its equivalents.
Does avoiding the word ‘means’ keep a limitation out of Section 112(f)?
No. After Williamson v. Citrix (2015), nonce words like module, unit, or mechanism paired with a function and no structure can invoke Section 112(f) even without the word means.
Why do software means-plus-function claims get invalidated?
Because the specification often fails to disclose an algorithm. Courts require a special-purpose computer programmed to perform a disclosed algorithm; a general-purpose computer alone leaves the claim indefinite.
How broad is a means-plus-function limitation?
Only as broad as the structure your specification discloses, plus equivalents. Disclosing several embodiments widens that scope; disclosing one narrows it dramatically.
Should I use functional claiming on purpose?
Sometimes. It can be useful when structure varies widely, but it should be a deliberate choice backed by solid structural disclosure, and ideally paired with parallel structural claims.