Patent Law Requirements: The Four Tests

“Patent law requirements” is not one test but four, and they come from three different sections of the U.S. patent statute. An invention disclosed to a university technology transfer office (TTO) has to clear all four before a patent can issue: the invention must be patent-eligible subject matter (35 U.S.C. § 101), novel (35 U.S.C. § 102), non-obvious (35 U.S.C. § 103), and adequately disclosed — enabled and supported by a written description (35 U.S.C. § 112). Each requirement has its own case law, its own failure modes, and its own dedicated guide on this site. This page is the map: what each requirement actually asks, how they differ from one another, where TTOs most often see inventions fail, and which guide to read next for the mechanics of any one of them.

The order below is also roughly the order a patent examiner works through them, and the order a TTO should think through a new invention disclosure: eligibility first (is this the kind of thing patent law covers at all), then novelty and non-obviousness together (is it new, and is it new enough), then disclosure (can the specification actually teach it).

1. Patent-eligible subject matter — 35 U.S.C. § 101

Section 101 is the threshold question, and it comes before novelty or obviousness are even relevant. It asks whether the invention falls into one of the statute’s eligible categories — a process, machine, manufacture, or composition of matter — and whether it avoids the judge-made exceptions for laws of nature, natural phenomena, and abstract ideas. The Supreme Court’s Mayo v. Prometheus (2012) and Alice Corp. v. CLS Bank (2014) decisions set out the two-step framework the USPTO now applies: first, is the claim directed to an ineligible concept; second, if so, do the claim elements add an “inventive concept” that amounts to significantly more than the abstract idea or natural law itself. This is where diagnostic-method inventions (a natural correlation between a biomarker and a disease state) and software inventions (an abstract idea implemented “on a computer”) most often run into trouble at university TTOs — an invention can be entirely novel and non-obvious and still fail § 101.

Full mechanics, the statutory text, and worked eligibility examples: 35 U.S.C. § 101: Patentable Subject Matter and the Alice/Mayo Eligibility Test.

2. Novelty — 35 U.S.C. § 102

Novelty asks a narrower question than eligibility: has this exact invention already been disclosed to the public, by anyone, before the relevant date? Under the America Invents Act’s first-inventor-to-file system, that date is the invention’s effective filing date, not the date of conception. Any single piece of prior art — a patent, publication, public use, sale, or public disclosure — that discloses every element of the claimed invention destroys novelty outright. The AIA also gives inventors a one-year grace period for their own disclosures, but that grace period is jurisdiction-specific (most of the rest of the world has no equivalent), which is exactly why a faculty member’s conference talk or preprint can quietly start a filing clock a lab isn’t tracking.

Grace-period mechanics and how they interact with a university’s internal disclosure timeline: 35 U.S.C. § 102: Patent Novelty and Invention Disclosure Timing.

3. Non-obviousness — 35 U.S.C. § 103

An invention can be technically novel — no single prior-art reference discloses it exactly — and still be unpatentable if the differences between it and the prior art would have been obvious to a person having ordinary skill in the relevant art at the time of filing. This is usually the hardest of the four requirements to argue and the one examiners rely on most often to reject claims, because it invites combining multiple prior-art references rather than requiring one reference to anticipate everything. The Supreme Court’s Graham v. John Deere (1966) factors (scope and content of the prior art, differences from the claims, the level of ordinary skill in the art, and any objective indicia of non-obviousness such as commercial success or long-felt need) still structure the analysis, and KSR v. Teleflex (2007) loosened the older, more rigid “teaching-suggestion-motivation” test the Federal Circuit had been applying, making obviousness rejections easier for examiners to support with common-sense reasoning about combining known elements.

The Graham factors and KSR’s effect in more depth: 35 U.S.C. § 103: Non-Obviousness, the Graham Factors, and KSR.

4. Enablement and written description — 35 U.S.C. § 112

The first three requirements are about whether the invention itself deserves a patent. Section 112 is about whether the application actually earns one — whether the specification discloses the invention clearly enough that a person of ordinary skill in the art could make and use the full scope of what’s claimed (enablement), and whether the specification shows the inventor actually possessed the claimed invention as of the filing date (written description). These are treated as two separate legal tests by the Federal Circuit even though both come from the same statutory sentence. Written description disputes come up most often when claims are amended or when a later application tries to claim priority back to an earlier filing; enablement disputes come up when claims are broader than what the specification actually teaches how to make and use. Section 112 also sets the format requirements for claims themselves (subsection (b)) and governs means-plus-function claiming (subsection (f)).

Statutory detail, the AIA’s best-mode change, and how this applies to provisional filings: 35 U.S.C. § 112: The Specification, Enablement, and Written Description Requirements.

How the four requirements interact

These aren’t independent checkboxes an examiner ticks in isolation — failing one can look like failing another, which is a common source of confusion when a TTO gets a rejection back. A few distinctions worth keeping straight:

Eligibility (§ 101) vs. novelty (§ 102). An invention can be entirely new (nobody has done it before) and still be ineligible subject matter (a natural correlation, an abstract idea). Novelty asks “has this been done,” eligibility asks “is this the kind of thing patent law covers at all.” A rejection citing § 101 is not cured by finding better prior art — it requires reframing the claim to add a concrete, technical application.
Novelty (§ 102) vs. non-obviousness (§ 103). Novelty requires one single reference to disclose every element of the claim; obviousness allows an examiner to combine multiple references and argue the combination would have been obvious. An invention can survive a novelty rejection (no single reference anticipates it) and still fail obviousness (a skilled person could have combined two known references to arrive at it).
Disclosure (§ 112) is orthogonal to the other three. An invention can be eligible, novel, and non-obvious and still fail if the specification doesn’t teach a skilled person how to actually make and use it, or claims more than the specification supports. This is why the drafting quality of a provisional application matters even though provisionals don’t require formal claims — a thin provisional can fail to support the priority claim a later non-provisional depends on.

Where this fits for a TTO evaluating a disclosure

In practice, a technology transfer office runs through a version of all four questions informally before deciding whether to file, well before a patent examiner ever sees the application: is this patentable subject matter at all (especially relevant for software and diagnostic-method disclosures), has it already been publicly disclosed in a way that starts or exceeds the grace-period clock, is it different enough from existing work to survive an obviousness combination, and can the inventor actually describe how to make and use it in enough technical detail to satisfy § 112. Getting any one of the four wrong is a common reason a patent application stalls in prosecution or, worse, issues and is later invalidated in litigation. For the mechanics of filing itself once these requirements are assessed, see the guides on provisional patent applications and the cost of filing a patent; for what happens after a patent issues, see patent licensing.

Frequently asked questions

What are the four requirements for patentability?

Patent-eligible subject matter (35 U.S.C. § 101), novelty (§ 102), non-obviousness (§ 103), and adequate disclosure through enablement and written description (§ 112). An invention must satisfy all four; failing any one is independently sufficient grounds for rejection or invalidation.

Is utility a separate requirement from the other four?

Utility is folded into § 101 itself — the statute requires a “new and useful” invention, so the eligibility analysis and the utility requirement share the same statutory sentence. In practice, utility rejections are rare for most university inventions (the bar is a specific, credible, substantial use) compared to the eligibility exceptions for abstract ideas and natural phenomena, which is why eligibility case law (Mayo/Alice) gets far more attention than utility case law.

Which requirement do university inventions fail most often?

There’s no single statistic that holds across all technology areas, but TTOs most commonly see software and diagnostic-method disclosures run into § 101 eligibility issues, and see obviousness (§ 103) rejections as the most frequent basis for an initial examiner rejection generally, since it allows combining multiple prior-art references rather than requiring one reference to match exactly.

Do these requirements apply the same way outside the United States?

The general structure — eligible subject matter, novelty, an inventive-step/non-obviousness analysis, and sufficient disclosure — is common across most patent systems because it derives from shared international norms, but the specific tests differ by jurisdiction. Novelty in particular is stricter in most of the rest of the world than in the U.S.: many jurisdictions have no equivalent to the AIA’s one-year grace period for an inventor’s own prior disclosure, so a public conference talk or preprint that’s still safe under U.S. § 102 can already have destroyed foreign filing rights. See the PCT patent application guide for how international filing timing interacts with this.

Patent Law Requirements: The Four Tests an Invention Must Pass