International patent law starts with the fact that international patents do not exist.
Every patent is national. Granted by one country's office, enforceable in that country only, subject to that country's law.
What the treaties do is coordinate filing — they preserve your priority date across borders and defer the moment you have to commit money country by country.
They grant nothing. That distinction is the whole subject.
The framework
| Instrument | Provides | Grants rights? |
|---|---|---|
| Paris Convention | 12-month priority period | No |
| PCT | One filing, national decisions deferred to ~30 months | No |
| TRIPS | Minimum standards for members | No |
| EPC / EPO | Centralised examination, then national validation | Grants, then fragments |
| Regional systems (ARIPO, OAPI, EAPO) | Regional filing routes | Varies |
| Hague | Registered design filing | Design registrations |
Only regional systems come close to granting anything supranational, and even the European patent fragments into national rights at grant.
The Paris Convention priority period
| Length | 12 months for utility inventions |
| Designs | 6 months |
| Runs from | Your first filing anywhere |
| Effect | Later foreign filings keep the original date |
| Extension | None |
This is the single most important international deadline. Missing it means foreign filings get their actual filing date, and anything published in between becomes prior art against them.
It runs from the first filing, not the last. Serial provisionals do not reset it. See patent priority date.
The PCT route
| Stage | Timing from priority |
|---|---|
| Priority filing | Day 0 |
| PCT application filed | Within 12 months |
| International search report | ~16 months |
| Publication | 18 months |
| Optional preliminary examination | ~22 months |
| National phase entry | ~30 months |
The PCT buys eighteen extra months to decide which countries justify the cost. That is its entire function and it is genuinely useful.
It does not examine for grant. The international search report is an indication, not a decision, and every national office examines independently afterwards.
National phase entry is where the money starts. Translations, local agents and official fees all apply per country.
Direct filing versus PCT
| Direct national filings | PCT then national phase | |
|---|---|---|
| Decision point | 12 months | ~30 months |
| Early cost | Higher | Lower |
| Total cost if many countries | Similar or lower | PCT fee added |
| Total cost if few countries | Lower | Higher |
| Information available at decision | Less | Search report, market data |
PCT wins where you are unsure. Direct filing wins where you already know you want two or three specific countries.
TRIPS sets a floor
| Provides | Detail |
|---|---|
| Minimum standards | For WTO members |
| Term | At least 20 years from filing |
| Non-discrimination | By field of technology |
| Does not | Grant anything, or harmonise procedure |
TRIPS explains why terms are broadly similar worldwide while procedures remain entirely national.
Europe fragments at grant
| Stage | What exists |
|---|---|
| EPO application | One application |
| EPO annuities | Payable from year 3, during pendency |
| Grant | One European patent |
| Validation | Per designated state, short window |
| After validation | Separate national rights |
| Renewal | National annuities, per state |
Validation is the real commitment, not the EPO filing. Validating in five states creates five renewal streams and five ways to lose a patent.
A family can be alive in three states and lapsed in four, and nothing centrally records that.
The Unified Patent Court allows centralised enforcement and revocation across participating states for patents that have not opted out — a significant development whose practice is still developing, so check current sources before relying on any description of it.
What the PCT actually delivers
| Deliverable | Value |
|---|---|
| International search report | Prior art found by a searching authority |
| Written opinion | Preliminary view on patentability |
| Publication at 18 months | Establishes prior art against others |
| Optional preliminary examination | A second opinion |
| Deferred national decisions | The main benefit |
| A granted patent | None — it grants nothing |
The search report is genuinely useful information arriving before you commit to national filings, and it frequently reshapes claim strategy.
No national office is bound by it. Each examines independently under its own law.
The grace period trap
| Jurisdiction | Grace period for your own disclosure |
|---|---|
| United States | 1 year |
| Most other countries | None |
| A few | Narrow, limited circumstances |
Public disclosure before filing forfeits rights in most of the world. Trade shows, crowdfunding pages, published papers, product sales and social posts all count.
The US grace period does not travel. Relying on it preserves US rights and destroys foreign ones in the same act.
This is the most common way foreign rights are lost, and it is entirely avoidable by filing first.
Priority claims must be made properly
| Requirement | Detail |
|---|---|
| Claimed in the ADS | Or the equivalent local form |
| Within the 12 months | No extension |
| Certified copy of the priority document | Frequently required |
| Omission | Earlier date at risk |
A priority claim that is not properly made does not exist, however genuine the earlier filing was.
National differences that change outcomes
| Issue | United States | Elsewhere |
|---|---|---|
| Grace period | 1 year | Usually none |
| Software eligibility | Alice two-step | EPO: technical character |
| Post-grant challenge | PGR, 9 months; IPR | EPO opposition, 9 months |
| Renewal | 3 fees after grant | Annual annuities |
| Payable during pendency | No | Often yes |
| Utility models | None | Available in several countries |
| Design term | 15 yrs from grant, no fees | Often 25 yrs with renewals |
Utility models are worth knowing about. Several countries offer a shorter-term, lightly examined right that can be faster and cheaper for structural inventions, with no US equivalent.
Utility models
| Feature | Detail |
|---|---|
| Available in | Several countries; not the US |
| Examination | Formality only, usually |
| Term | Typically shorter — often 10 years |
| Grant speed | Much faster |
| Subject matter | Often limited to shape or structure |
| Prior art value | Full — they publish |
Utility models are worth considering for structural inventions in countries that offer them, since they grant quickly and cheaply.
They also matter as prior art, and they are the category most often missed in searching. See Chinese patentability search.
Cost escalates abroad
| Cost element | Detail |
|---|---|
| PCT filing | One-off |
| National phase entry, per country | Official fees plus local agent |
| Translation | Substantial in non-English jurisdictions |
| Prosecution per country | Independent |
| Annuities | Annual, escalating, per country |
| Validation, Europe | Per state |
| US | Most other jurisdictions | |
|---|---|---|
| Renewal events | 3 | Annual — up to 20 |
| Decisions per patent | 3 | Up to 20 |
A four-country family generates roughly twenty deadlines a year, which is beyond reliable manual tracking. See patent annuity fees by country.
Translation costs
| Jurisdiction | Translation |
|---|---|
| Japan, China, Korea | Full translation — substantial |
| Europe, at EPO stage | One of three official languages |
| Europe, at validation | Varies by state; some require full translation |
| Latin America | Full translation |
| English-language jurisdictions | None |
Translation is frequently the largest single line in a national phase entry, and it scales with specification length.
A long specification costs more in every country. That is a real argument for concision where it does not sacrifice support.
The PCT is not a patent
| Misconception | Reality |
|---|---|
| "I have a PCT patent" | The PCT grants nothing |
| "It covers all member states" | It preserves the option to file |
| "Examination is done" | Each office examines independently |
| "Protection has started" | No enforceable rights |
Choosing countries
| Ask | Why |
|---|---|
| Where will we sell? | Revenue justifies cost |
| Where do competitors manufacture? | Reaches the supply chain |
| Where would we actually enforce? | Practical reality |
| Is the market growing? | Forward view |
| What is the escalating annuity curve? | Later years cost more |
Manufacturing coverage is undervalued. A patent where a product is made can stop production rather than chasing distribution.
Importation is the fallback. A US patent reaches goods made abroad at the point they enter the country, which is the answer where local filing is not affordable.
Claims differ by country
| Why | Detail |
|---|---|
| Independent examination | Each office applies its own law |
| Different prior art cited | Local databases and languages |
| Different eligibility standards | Especially software |
| Different amendment practice | What is permitted varies |
| Result | Same family, different scope per country |
Never assume the German claims match the US ones. They are frequently narrower, broader, or aimed at different aspects entirely.
Read the claims of each member before relying on a family for anything. See patent family.
Worked example: one invention, four routes
A US applicant with a mechanical invention, priority filed March 2025.
| Route | Decision point | Countries | Cost profile |
|---|---|---|---|
| US only | — | 1 | Lowest |
| Direct Paris filings | Mar 2026 | US, EP, CN | Early commitment |
| PCT then national phase | Sep 2027 | Decided later | PCT fee, deferred spend |
| PCT, then abandon abroad | Sep 2027 | US only | PCT fee wasted |
What the extra eighteen months bought
| By Sep 2027 the applicant knew | Effect |
|---|---|
| Which markets had traction | Dropped one planned country |
| The international search report | Adjusted claim strategy |
| Whether a competitor had emerged | Added a manufacturing jurisdiction |
| Actual revenue | Justified the spend |
The PCT fee bought better information at the decision point. That is what it is for.
Direct filing would have committed the money in March 2026 on assumptions rather than data.
Foreign filing licences
| Requirement | Inventions made in the US generally need one before filing abroad |
| Normally granted | On the filing receipt |
| Filing abroad first without one | Can invalidate a later US patent |
| Retroactive licence | Available by petition in some circumstances |
| Other countries | Similar rules exist elsewhere |
Check the filing receipt before filing abroad. The licence is usually granted automatically and confirming it costs nothing.
Several other countries have equivalent rules for inventions made within their territory, which matters for distributed R&D teams.
Where costs land
| Stage | Cost weight |
|---|---|
| Priority filing | Low |
| PCT filing | Moderate |
| National phase entry | Highest single step |
| Prosecution per country | Ongoing |
| Validation, Europe | Per state |
| Annuities | Escalating, forever |
National phase entry is the decision point that matters financially, which is precisely why the PCT defers it.
Enforcement is national too
| A US patent is enforced | In US courts |
| A German patent | In German courts |
| Cross-border disputes | Parallel proceedings |
| Judgments | Do not automatically travel |
| UPC | Centralised for participating European states |
Multi-country disputes mean multi-country litigation, with different procedures, timelines and outcomes in each.
Regional systems
| System | Covers |
|---|---|
| EPO | European states, then national validation |
| ARIPO | Participating African states |
| OAPI | Members share a single unitary right |
| EAPO | Eurasian states |
OAPI is unusual in producing a single right across its members rather than fragmenting into national patents, which makes it administratively simpler than the European route.
What the outcome usually is
| Ipiry Patent Survival Curve v1.0 | Rate |
|---|---|
| Survive the 3.5-year fee (2022 cohort) | 85.8% |
| Survive the 7.5-year fee (2018 cohort) | 64.6% |
| Reach full term (2014 cohort) | 41.4% |
| Abandoned before full term | 58.6% |
Computed from 27,273,654 USPTO maintenance fee records covering 8,262,336 US utility patents — see the patent survival curve.
That is the domestic rate with only three decisions. Foreign members face the question annually and are typically pruned earlier.
Common international mistakes
| Mistake | Consequence |
|---|---|
| Disclosing before filing | Foreign rights forfeited |
| Missing the 12-month deadline | Priority lost |
| Filing everywhere reflexively | Escalating annuities on unused rights |
| Assuming US claims apply abroad | They differ |
| Forgetting validation deadlines | European rights lapse |
| No annuity service for a foreign portfolio | Missed deadlines |
International patent law: the checklist
- There is no world patent. Rights are national, always.
- Diarise twelve months from your first filing for all foreign applications.
- Never disclose publicly before filing. Most countries have no grace period.
- Use the PCT where you are unsure, direct filing where you already know.
- Treat European validation as the real commitment, not the EPO filing.
- File where you sell and where competitors manufacture.
- Budget annuities, not just filing. Annual and escalating abroad.
- Expect different claims to issue in different countries.
- Remember importation reaches goods made where you hold nothing.
- Check current sources on the UPC. Its practice is still developing.