§103 · Non-obviousness
The bar that rejects the most patents.
Novelty asks whether your invention is new. Non-obviousness (§103) asks the harder question: even if no single reference shows it, would it have been obvious to combine what was already known? It is the single most common reason applications are rejected — here is exactly how the test works.
The framework
The four Graham factors
Every obviousness analysis since Graham v. John Deere (1966) rests on four factual questions. Get these straight and the rest follows.
- 01
Scope and content of the prior art
Determine everything that was already known in the field — the patents, publications, and products a skilled person would have had in front of them at the time of filing.
- 02
Differences between the prior art and the claim
Pin down exactly what the claim adds over the closest prior art. Obviousness turns on this gap, not on the invention as a whole.
- 03
Level of ordinary skill in the art
Define the hypothetical 'person of ordinary skill' — their education, experience, and ordinary creativity. A higher skill level makes more things obvious.
- 04
Secondary considerations
Weigh real-world, objective evidence — commercial success, long-felt need, failure of others — that can show an invention was not obvious after all.
How examiners argue it
The KSR rationales
In KSR v. Teleflex (2007) the Supreme Court loosened the old rigid test. An examiner can now find a claim obvious through any of these recognized rationales — not just an explicit teaching to combine.
Combining known elements
Combining prior-art elements by known methods, where each works as expected, to yield predictable results.
e.g. Bolting a known sensor onto a known machine to get the reading everyone would expect.
Simple substitution
Substituting one known element for another known equivalent to obtain predictable results.
e.g. Swapping a known steel part for a known aluminum one to save weight.
Use of a known technique
Using a known technique to improve a similar device, method, or product in the same way it has improved others.
e.g. Applying a standard anti-corrosion coating used elsewhere to a new but similar component.
Applying a known technique to a ready device
Applying a known technique to a known device or method that is ready for the improvement, with predictable results.
e.g. Adding a well-known feedback loop to a controller that obviously needed one.
Obvious to try
Choosing from a finite number of identified, predictable solutions with a reasonable expectation of success.
e.g. Testing the handful of known catalysts for a reaction when one was likely to work.
Design incentives & market forces
Known work in one field prompting variations in the same or a different field based on design incentives or market forces.
e.g. Adapting a proven phone-hinge design to a tablet because the market pushed that way.
Teaching, suggestion, or motivation
An explicit teaching, suggestion, or motivation in the prior art to combine the references. Still valid after KSR — but no longer the only path to obviousness.
e.g. A reference that literally says 'this part could be combined with a device like X.'
How you fight back
Secondary considerations
These are your evidence. Real-world proof that an invention was notobvious can overcome even a strong prima facie rejection — as long as you can tie it to the patented feature (a “nexus”).
Commercial success
The product sold well because of the patented feature itself — not marketing or an unrelated advantage (a 'nexus' is required).
Long-felt but unmet need
The problem had persisted for years and the industry wanted a solution that no one had delivered.
Failure of others
Skilled people actively tried to solve the problem and failed before this invention.
Unexpected results
The invention works markedly better, or differently, than a skilled person would have predicted.
Copying
Competitors copied the invention instead of using the prior art or designing around it.
Industry praise or skepticism
Experts doubted the approach would work, or praised it as a breakthrough once it did.
Licensing and acquiescence
Others took licenses to the patent, tacitly acknowledging it covers something non-obvious and valuable.
When the rejection lands
How to respond to a §103 rejection
Most applications get at least one obviousness rejection — it is a normal part of prosecution, not a death sentence. Here is the response sequence, ordered cheapest-fix-first: win on argument before you amend, and amend before you appeal.
Map exactly what the examiner asserted.
A §103 rejection names one or more references and a reason to combine them. Before responding, chart it: which reference is cited for which claim limitation, and which rationale (see above) carries the combination. You cannot rebut an argument you haven't pinned down element by element.
Look for a missing claim element first.
The cleanest win needs no amendment: if no cited reference — alone or combined — discloses a specific limitation in your claim, the prima facie case fails. Identify the limitation the art simply doesn't teach and say so plainly. This is the most common successful §103 response.
Attack the reason to combine.
KSR still requires the examiner to articulate a reason with rational underpinning — a conclusory “it would have been obvious” is rebuttable. Argue there was no motivation to combine, that combining the references would change a reference's principle of operation or render it unsuitable for its purpose, or that the art teaches away from the combination.
Call out impermissible hindsight.
Examiners can only use knowledge available before your filing date. If the rejection reads like it used your own application as a roadmap to assemble scattered pieces of prior art, that is hindsight bias — name it, and tie your argument to what a skilled person actually knew at the time.
Submit secondary-considerations evidence — with a nexus.
Commercial success, long-felt but unmet need, failure of others, unexpected results, industry praise, copying, or skepticism can overcome even a strong rejection — but only if tied to the claimed feature (the nexus). This usually goes in as a declaration under 37 CFR 1.132 from an inventor or expert.
Amend only what the specification supports.
If argument alone won't carry it, narrow a claim to a limitation the cited art neither shows nor suggests — but only to subject matter your specification already disclosed (no new matter). Remember that every narrowing amendment can create prosecution-history estoppel, shrinking what you can later assert against infringers.
Request an examiner interview.
A 30-minute call often resolves a §103 rejection faster than rounds of written argument — you can walk the examiner through the missing element or agree on allowable claim language directly. Interviews are free and frequently the highest-leverage move.
If the rejection is final: RCE, after-final, or appeal.
A final rejection narrows your options to a Request for Continued Examination (reopens prosecution), the after-final pilot programs, or an appeal to the PTAB when you believe the examiner is simply wrong on the law. Each has its own cost and timeline — see the office-action guide for the full decision tree.
Check your idea
Non-obviousness is one of four gates.
An invention must also be eligible (§101), novel (§102), and fully described (§112). Run yours through all four, or start with a prior-art search to see what you're up against.
Plain-English education, not legal advice. Obviousness is intensely fact-specific and hindsight is a constant trap — for a real assessment of your claims, work with a registered patent attorney. PatentBrief is not a law firm.
FAQ
Non-obviousness — common questions
What does non-obviousness mean in patent law?
Under §103, an invention is unpatentable if it would have been obvious to a person of ordinary skill in the field at the time of filing — even if no single prior-art reference shows the whole thing. It's the most common reason applications are rejected.
What are the Graham factors?
From Graham v. John Deere (1966), the four factual inquiries behind every obviousnessobviousnessA 103 rejection: the invention would have been obvious to a skilled person who combined existing prior art. The most common rejection in patent prosecution.Read more → analysis: the scope and content of the prior artprior artEarlier patents, publications, or products that existed before this patent's filing date. Patent claims must be novel over the prior art.Read more →, the differences between the prior art and the claimclaimA numbered sentence at the end of a patent that legally defines what the inventor owns. The most important section.Read more →, the level of ordinary skill in the art, and secondary considerations.
What is the 'obvious to try' rationale?
From KSR v. Teleflex (2007): a claimclaimA numbered sentence at the end of a patent that legally defines what the inventor owns. The most important section.Read more → can be obvious if it was simply a matter of choosing from a finite number of identified, predictable solutions with a reasonable expectation of success — you don't need an explicit teaching to combine references.
The four requirements