Direct comparison
Trade Secrets vs. Patents: IP Strategy
Compare patent and trade secret protection: disclosure requirements, term length, what each actually stops competitors from doing, and Bayh-Dole implications.
Side-by-side comparison
| Dimension | Patent | Trade Secret |
|---|---|---|
| Legal basis | 35 U.S.C. (the Patent Act), administered by the USPTO. A utility patent requires the invention to be eligible subject matter (35 U.S.C. 101), novel (102), non-obvious (103), and disclosed with enough detail to enable someone skilled in the field to practice it (112). | Federal Defend Trade Secrets Act of 2016 (18 U.S.C. 1839(3)) and the state-level Uniform Trade Secrets Act, adopted in some form by 49 states and DC (New York relies on common-law doctrine instead). Both define a trade secret the same way: information with independent economic value from not being generally known or readily ascertainable, subject to reasonable efforts by the owner to keep it secret. |
| Disclosure requirement | Mandatory. A U.S. patent application publishes 18 months after the earliest filing date (with limited exceptions), and the granted patent's specification is a permanent public document describing the invention in enabling detail. | None. No filing, no registration, no public document. The information simply has to actually be kept secret -- confidentiality agreements, access controls, and similar 'reasonable measures' are how an owner satisfies the legal test, not a formality layered on top of it. |
| Duration of protection | Fixed term: 20 years from the earliest non-provisional filing date under 35 U.S.C. 154(a)(2) (design patents run 15 years from grant instead, under 35 U.S.C. 173). Maintenance fees are due at 3.5, 7.5, and 11.5 years after grant; missing one lets the patent lapse early. A provisional application does not shorten the 20-year term -- the clock starts at the non-provisional filing. | No statutory limit. Protection continues indefinitely as long as the information stays secret, has economic value from that secrecy, and the owner keeps taking reasonable measures to protect it. It ends the moment any one of those three conditions fails -- most often because the secret is disclosed, published, or independently developed by someone else. |
| What it actually stops others from doing | Making, using, selling, offering to sell, or importing the claimed invention in the country of grant, full stop -- including a competitor who reinvents the exact same thing entirely independently, with no knowledge the patent existed. | Only acquisition through improper means (theft, breach of a confidentiality obligation, industrial espionage, bribery, and similar conduct as defined under the DTSA/UTSA). It does not stop a competitor who reaches the same result through independent research and development, or who lawfully reverse-engineers a product built on the secret -- both are explicitly carved out of 'improper means' under both statutes. |
| Ongoing obligations to keep protection | Pay maintenance fees on schedule; no ongoing secrecy or use requirement (the invention can be widely marketed and still stay protected for the full term). | Continuous 'reasonable measures' to preserve secrecy -- NDAs with employees, contractors, and partners; restricted physical and digital access; marking and segregating confidential material; and documented security practices. Courts have found trade secret status forfeited where an owner's own conduct failed to satisfy this element, independent of what a defendant did. |
| Cost profile | Front-loaded and largely fixed: USPTO filing/examination/issue fees, typically substantial attorney costs to draft and prosecute the application, plus maintenance fees over the life of the patent. Costs are incurred whether or not the invention ever generates revenue. | No filing or government fees at all. Cost is ongoing and operational -- the systems, agreements, and internal controls needed to actually maintain secrecy -- and scales with how long and how widely the information needs to be protected. |
| Enforcement | Infringement suit in federal court (or a USPTO post-grant proceeding) against the accused product or process itself; the patent owner does not have to prove the defendant copied anything or acted improperly -- independent invention is not a defense. | Misappropriation suit under the DTSA (federal) or state UTSA, which requires proving both that the information met the legal definition of a trade secret AND that the defendant acquired, used, or disclosed it through improper means -- a materially higher evidentiary burden than patent infringement, and one that depends on the strength of the owner's own secrecy practices. |
| Best fit | Inventions that are easy to reverse-engineer once sold or deployed (a physical product, a chemical composition, a device), where the whole value of protection is stopping others from replicating something the market can see. | Know-how that a competitor genuinely could not reconstruct just by examining the end product -- process parameters, internal formulations, algorithms that never leave a server, negative know-how ('we tried these approaches and they don't work'), and anything where the underlying research would take a rival years to independently replicate. |
Common questions
FAQ
Can the same invention be protected by both a patent and a trade secret?+
Not the same information, and not at the same time. Filing a patent application requires disclosing the invention publicly (via the 18-month publication and the eventual granted patent), which destroys trade-secret status for whatever is disclosed. Some organizations do use a layered strategy -- patenting the core invention while keeping surrounding process know-how (manufacturing parameters, optimization details not required for an enabling disclosure) as a trade secret -- but that is two different bodies of information under two different regimes, not dual protection of the same secret.
Once a trade secret decision is made, can an inventor change their mind and file a patent later?+
Only if the information genuinely hasn't been disclosed and hasn't been on sale or in public use in a way that starts a statutory bar clock under 35 U.S.C. 102. In practice, the longer something has operated as a trade secret -- especially if it has been sold as part of a commercial product -- the more likely a later patent application runs into a novelty or on-sale bar problem. This is why the decision is usually made deliberately and early, not left open.
Does choosing trade secret protection avoid a university's Bayh-Dole obligations for federally funded research?+
No. Under the Bayh-Dole Act's implementing regulations (37 CFR 401.14), an institution must disclose each subject invention made under a federal award to the funding agency within a set window regardless of whether it intends to seek patent protection. Deciding not to patent an invention is a legitimate choice, but it does not remove the underlying obligation to report the invention to the sponsoring agency in the first place; it only changes what happens after that disclosure.
Going deeper
