A patent annuity is a renewal fee paid every year to keep a patent alive in a country outside the United States.

The US is the outlier. Three maintenance fees after grant, nothing while the application is pending. Most of the world charges annually, often from year three, and often before anything has been granted.

The difference compounds. A US patent costs $14,470 in maintenance fees across its life at large entity rates. A four-country family can exceed that every few years, and the annual bill rises each time.

Where annuities apply

Nearly everywhere except the United States.

Jurisdiction Renewal model Starts
United States 3 maintenance fees 3.5 years after grant
European Patent Office Annual, then national after grant Year 3 from filing
United Kingdom Annual Year 5
Germany Annual Year 3
Japan Annual From grant
China Annual Year 1 from filing
Canada Annual Year 2
Australia Annual Year 5

Several charge from year one or two of filing, long before grant. An application pending for six years in China or Germany has already generated several annuity payments by the time anything is enforceable.

The EPO case is the one that confuses US applicants most. Renewal fees are paid centrally to the EPO while the application is pending, and then the granted European patent becomes a bundle of national patents, each with its own annuity to its own office.

The Unitary Patent is the exception. It carries a single renewal fee covering participating states, which removes the bundle problem for those countries but requires the choice to be made at grant.

Annuities versus US maintenance fees

US maintenance fees Annuities
Frequency Three payments Every year
Timing 3.5, 7.5, 11.5 years after grant Typically from year 3 after filing
Payable while pending No Often yes
Escalation Three steps Rises annually
Paid to One office Each national office separately
Design patents None due Often payable
Grace period 6 months with surcharge Commonly ~6 months, varies

Payability during pendency is the structural difference that costs the most. An application examined for six years at the EPO generates six years of renewal fees before any enforceable right exists.

And they are per country. A European patent that has granted becomes a bundle of national patents, each needing its own annuity paid to its own office through its own agent.

Why they escalate

By design. Rising fees force an annual re-evaluation, so patents that stop earning their keep are released into the public domain rather than held indefinitely.

It works, and the US data shows how well.

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%
Lost at the third fee 23.2 points

Computed from 27,273,654 USPTO maintenance fee records covering 8,262,336 US utility patents — see the patent survival curve.

Attrition is higher abroad than these US figures suggest. With a decision forced every year rather than three times, and the cost rising each time, foreign family members are typically dropped earlier than their US counterparts.

Which means the shape of a family changes over time. A patent filed in eight countries at year three is often maintained in two by year fifteen, and that is normal portfolio management rather than failure.

Restoration when an annuity is missed

A lapse is sometimes recoverable, and the standard varies sharply by country.

Standard What it requires Where it is common
Unintentional The failure was not deliberate US-style, more forgiving
Due care You had systems in place and they failed despite reasonable care European standard, much harder
Automatic grace period Payment plus surcharge within a set window Most jurisdictions, ~6 months
No restoration Nothing available after grace Some jurisdictions

The due care standard defeats most applications. Forgetting is not due care. An organisation with no docketing system, no reminders and no backup has difficulty showing the failure happened despite reasonable precautions.

Third-party intervening rights complicate recovery. Someone who began practising the invention while the patent was lapsed may keep doing so even after restoration, which turns a recovered patent into a partially unenforceable one.

Restoration windows are short and evidence-heavy. The practical answer is a duplicate calendar and a payment service with insurance, because the cost of both is far below the cost of one contested restoration.

The cost trajectory

Early years are cheap. Late years are not.

Stage Typical annual cost per country What it feels like
Years 3–5 Low hundreds Barely noticed
Years 6–10 Several hundred Noticeable across a family
Years 11–15 Substantial Prompts serious pruning
Years 16–20 Highest Only clear commercial value survives

Multiply by countries, then add agent fees. Official fees are only part of the cost — local agent charges, currency handling and, in some jurisdictions, translation requirements all sit on top.

The compounding is what catches people. A family of four patents in five countries is twenty separate annuity payments a year, each rising annually, each with its own deadline in its own jurisdiction.

Worked example: a four-country family

One invention, filed in the US, Europe, Japan and China.

Phase US Europe Japan China
Filing to grant No fees due Annuities from year 3 Fees from grant Annuities from year 3
Grant year Year 4 Year 6 Year 5 Year 5
Payment pattern 3 fees after grant Annual, then per country after grant Annual Annual
Fees paid by year 10 2 8 6 8
Fees paid by year 20 3 18 16 18

Three payments against eighteen. That is the structural difference in one table, and it is why international families are pruned far more aggressively than US-only holdings.

The year 11 decision

By year 11 the family costs are rising sharply and the owner reviews it.

Jurisdiction Commercial position Decision
US Largest market, product shipping Keep
Germany (from the EP bundle) Manufacturing partner located there Keep
France, Italy, Spain (from the EP bundle) No sales, no competitors Drop
Japan Licensee negotiation ongoing Keep, review in 12 months
China Manufacturing risk, no sales Keep — defensive

Dropping three European countries removes three annuities a year, permanently, and rising. The saving compounds for the remaining nine years.

Nothing is recoverable once dropped. A lapsed national patent cannot be revived except in narrow circumstances, so pruning is a one-way decision and deserves the review it forces.

Annuities on applications you may abandon

Paying renewal fees on a pending application is money spent before any right exists, and the decision deserves the same scrutiny as a granted patent.

Question at each pending-application annuity If no
Is prosecution progressing toward allowable claims? Consider abandoning now
Would the likely allowed scope be commercially useful? Reconsider
Do we still sell or manufacture in this market? Drop the country
Is a competitor active here? Defensive case may still hold

A long, contested prosecution in a market you have left is the clearest waste in most portfolios. The fees continue annually while the eventual right may be both narrow and irrelevant.

Abandoning a pending application is cheaper than abandoning a granted patent, because nothing has been spent on grant formalities and no term has been lost that mattered.

Managing annuities

Approach Suits Risk
In-house tracking 1–3 patents, one or two countries Missed deadline ends the patent
Annuity payment service Portfolios beyond a handful Service fees; usually insured
Law firm handling Where firms already manage prosecution Highest cost per payment

Payment services earn their fee on scale. Tracking twenty deadlines across five jurisdictions, in five currencies, through five local agents, is exactly the kind of administration where one lapse costs more than years of service fees.

Ask what insurance they carry. Reputable services carry cover against missed payments, and that cover is a substantial part of what you are buying.

Keep your own calendar regardless. Outsourcing the payment does not outsource the consequence, and a duplicate diary of deadlines is cheap insurance.

Instruct on the decision, not just the payment. A service will pay whatever you have not told it to stop paying. The annual review is yours.

Pruning before the cost compounds

The annual cadence is a feature. It forces a decision every year, and taken seriously it keeps a portfolio honest.

Question at each renewal If no
Are we selling in this market? Consider dropping
Is a competitor manufacturing here? Defensive value may justify keeping
Is there a licensee or negotiation live? Keep pending outcome
Would a buyer want this national right? Consider selling before lapse
Does remaining term justify the escalating fee? Late-life patents rarely do

Selling before lapse is the option most owners skip. A national patent that still has commercial value to someone else is worth something up to the deadline and nothing afterwards.

Review the whole family at once, once a year. Country-by-country decisions made in isolation as each deadline arrives produce incoherent portfolios — the wrong markets kept because their deadline fell in a quiet month.

See patent portfolio pruning for the review process and patent maintenance fees for the US schedule the annuity system replaces elsewhere.

Patent annuity: the checklist

  1. Map every deadline across every jurisdiction in one calendar, including pending applications where annuities are already running.
  2. Budget for escalation. This year's total is not next year's; annuities rise annually and the increases accelerate.
  3. Confirm which countries charge on pending applications. European renewal fees start at year three regardless of grant.
  4. Remember a granted European patent becomes national patents, each with its own annuity, agent and deadline.
  5. Review the whole family once a year, not country by country as each deadline arrives.
  6. Drop markets where you do not sell and competitors do not manufacture. That is normal management, not failure.
  7. Consider selling national rights before lapsing them. They are worth something until the deadline and nothing after.
  8. Use a payment service beyond a handful of patents, and ask specifically what insurance it carries.
  9. Keep an independent calendar even when a service handles payment. The consequence of a lapse is yours.
  10. Weigh escalating fees against remaining term. Late-life annuities in marginal markets are the clearest case for release.