A patent independent claim is a single sentence that defines, on its own, what the patent covers. It does not refer back to any other claim. It recites every element the invention requires, and because it carries no incorporated limitations from elsewhere, it is the broadest claim in the patent and the one that marks the outer edge of protection. When practitioners say "read the independent claim," they mean exactly this: ignore the abstract, ignore the marketing, and look at the self-contained claim that lists what must be present for something to fall within the patent's scope.

The form of a claim is fixed by statute. Under 35 U.S.C. 112, the specification must conclude with the claims, and the statute then describes the permitted forms. Subsection (b) requires that the specification "conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the inventor or a joint inventor regards as the invention," and subsection (c) sets out the available structures.

A claim may be written in independent or, if the nature of the case admits, in dependent or multiple dependent form. ... A claim in dependent form shall contain a reference to a claim previously set forth and then specify a further limitation of the subject matter claimed. A claim in dependent form shall be construed to incorporate by reference all the limitations of the claim to which reference is made.— 35 U.S.C. 112(c)–(d), source

That statutory text is the whole distinction in miniature. A dependent claim must "contain a reference to a claim previously set forth" and add "a further limitation," and it is "construed to incorporate by reference all the limitations of the claim to which reference is made." An independent claim does none of that. It contains no back-reference and incorporates nothing from another claim, so everything it covers is written into its own text. The practical consequence follows directly from the incorporation rule: because a dependent claim carries every limitation of its parent plus at least one more, it is by definition narrower than the claim it depends on. The independent claim it traces back to is therefore the broadest member of that claim family.

What the rules add: order, antecedent basis, and counting

The implementing regulation, 37 CFR 1.75, fills in the mechanics. It restates that "the specification must conclude with a claim particularly pointing out and distinctly claiming the subject matter which the applicant regards as his invention," and that "one or more claims may be presented in dependent form, referring back to and further limiting another claim or claims in the same application." The rule also handles a complication: a multiple dependent claim, which refers to more than one other claim, "shall refer to such other claims in the alternative only," and for fee purposes is counted as the number of claims to which it refers. That counting rule is why applicants pay attention to how many independent claims they file — fees scale with claim count, and each independent claim is its own anchor.

The rule also describes a recommended structure for an independent claim in an improvement case. Per 37 CFR 1.75(e), where the nature of the case admits, an independent claim "should contain in the following order" a preamble describing the conventional or known elements, a transitional phrase "such as 'wherein the improvement comprises,'" and then the elements that constitute "the new or improved portion." This is the classic Jepson form. Most modern claims do not use it, but the rule illustrates the anatomy every independent claim shares: a preamble, a transition, and a body of elements. Each element in that body is a limitation, and for an accused product or method to infringe, it must contain every one of those limitations — a property courts call the all-elements rule. Add a limitation to the claim and you shrink its scope; remove one and you widen it. That is why the independent claim, carrying the fewest limitations in its family, covers the most.

Reading an independent claim in the wild

A real example makes the structure concrete. U.S. Patent US11533175B1, "Systems and methods for post-quantum cryptography on a smartcard," assigned to Wells Fargo Bank, N.A. and granted December 20, 2022 from an application with a January 30, 2020 priority date, opens with this independent claim 1: "A system for post-quantum cryptography (PQC) comprising a PQC smartcard, wherein the PQC smartcard comprises: PQC cryptographic algorithm selection circuitry that selects a PQC cryptographic technique from a set of PQC cryptographic techniques for encrypting the data; and PQC cryptographic circuitry that encrypts data based on a generated set of PQC encryption attributes and the PQC cryptographic technique." Notice that the claim refers to no other claim. It names a smartcard, two circuits, and what each circuit does, and it stops. Everything required to be within this claim is present in its own words — that is what makes it independent. The dependent claims that follow in the same patent each open with a reference such as "The system of claim 1, wherein..." and add a narrowing detail; under 35 U.S.C. 112(d) each of those carries all of claim 1's limitations plus its own.

Two cautions follow from the statute. First, an independent claim defines coverage, not a product: the existence of claim 1 above means Wells Fargo claimed a smartcard architecture, not that any particular card ships. Filing is not shipping, and a claim's words describe a legal boundary, not a feature list. Second, the broadest claim is not always the strongest. A broad independent claim covers more, but it is also more exposed to prior art that an examiner or a challenger can read onto its few limitations; narrower dependent claims often survive when a broad parent falls. For anyone trying to understand what a security or cryptography patent actually protects, the method is the same every time: find the independent claims, read each limitation, and treat that self-contained text — not the title and not the abstract — as the measure of the patent's reach.