How Many Were Printed? What Banknote Print Runs Really Measure

“How many were printed?” sounds like one of the simplest questions a banknote collector can ask. It is often one of the hardest to answer correctly. A figure attached to a note may describe a forecast, an authorised ceiling, an order, a contract, completed production, delivery, release into circulation, the stock still outstanding, notes destroyed, or examples submitted to a grading company. Those figures can all be accurate and still count different things.

This guide explains how to tell them apart. It combines official issuer records with a structured review of 225 country chapters of The Banknote Book, covering 11,075 pages and about 3.73 million words. The result is not a global table of print runs. It is a method for deciding what a published number actually measures before using it in a catalogue, collection record or sales description.

A number is not a print run until its source says what was counted, for which note, over which period, and at which stage between planning and destruction.

Why “How Many Were Printed?” Is Harder Than It Sounds

A banknote does not move directly from an idea to a wallet. An issuing authority forecasts demand, approves a programme, assigns or contracts production, receives finished stock, releases notes when needed, withdraws worn examples and eventually destroys them. Different institutions may control different stages. The organisation placing an order may not be the printer, and the quantity released in one year may include notes printed in an earlier one.

The language used in records is just as important as the number. Planned, ordered, contracted, produced, delivered, issued and outstanding are not interchangeable verbs. A range such as 4.8–5.7 billion describes uncertainty or permitted adjustment; “around 601 million” is a rounded figure; and a stock measured on one date cannot be compared directly with a flow accumulated through a year.

This is separate from the physical process described in How Banknotes Are Printed. Printing techniques explain how the object was made. A production figure explains how many objects a source says reached a defined stage. Before using one, the collector must identify both the stage and the scope: denomination, series or variety; national issue or export order; single printer or whole system; month, fiscal year, calendar year or point-in-time stock.

Eleven Different Numbers Can Sit Behind One Question

A practical source audit separates eleven measurements. Contractual delivery is shown within the contract stage because it answers a different question from the amount originally agreed. Only one row below is unambiguously completed physical production, and even that figure still needs a denomination, period and source.

MeasurementWhat the figure actually countsWhy it is not automatically a print run
1. Forecast demandExpected future needA planning input; no order or production has necessarily occurred
2. Authorised ceilingThe maximum approved quantityPermission is not use of that permission
3. OrderWhat an issuer asks a printer to supplyOrders can be revised, ranged or shifted between denominations
4. AllocationA production share assigned to a printer or central bankAllocation precedes completed output
5. Contract and contractual deliveryWhat parties agreed, and what was later delivered under that agreementThe two figures may differ and neither proves issue into circulation
6. Produced or printedFinished physical notes reported by the producer or issuerThis is the closest strict answer, but it still may aggregate several notes or customers
7. Delivered to issuerNotes transferred from printer to customerDelivered stock can remain in vaults
8. Issued or releasedNotes put into the monetary systemReleased notes may have been printed in another period
9. OutstandingNotes in circulation at a date or annual averageThis is a changing stock after issue, return and destruction
10. Returned, unfit or destroyedNotes removed from the usable populationA withdrawal flow cannot reconstruct original production on its own
11. Surviving or graded populationKnown survivors or notes certified by a grading serviceUnknown and unsubmitted notes remain invisible

The important discipline is to preserve the source’s own category. If an annual report says delivered under contract, that is the claim. Turning it into printed adds a production conclusion that the source did not state. Turning it into issued adds a monetary conclusion as well.

An Order Is Not Production: The United States

The United States provides an unusually clear demonstration because the order and the production result are published by different parts of the system. The Federal Reserve Board publishes the Federal Reserve note print order. The Bureau of Engraving and Printing publishes separate annual production reports stating that their figures represent notes printed.

The Board’s order approved in July 2026 for calendar year 2027 is a range of 4.8 to 5.7 billion notes, valued at approximately $132.1 to $166.4 billion. It is based on estimated net demand, inventory, destruction of unfit notes and other operational needs, and the Board says denomination-level production may be adjusted during the year. The range is therefore a plan with room to change, not a completed count.

Period labels create another trap. The Federal Reserve’s historical note says years before 2024 use the federal fiscal year, while 2024 onward use the calendar year. BEP’s annual table remains organised by federal fiscal year; for example, it reports 2,397,104,000 one-dollar notes printed in FY2023. The order and production series can be compared only after their periods and definitions are aligned. Copying two official numbers does not make them comparable.

The strongest evidence is not the largest number. It is the number with the clearest verb, unit, scope and period.

A Contract Is Not a Print Run: Brazil’s Five Decision Stages

An official Brazilian record shows how several large quantities can accumulate around one procurement history without any of them becoming a verified production total. Minutes of the National Monetary Council from July 1967 retrospectively describe contracts made on 14 April 1964 with the American Bank Note Company and Thomas De La Rue. Each supplier had been contracted for 600 million notes: 1.2 billion in aggregate.

The same sequence then records a later need for another 600 million notes; an order approved on 7 March 1967 for 300 million from each supplier; a proposed revision to 200 million notes of the NCr$10 denomination; and alternatives of 130 million or 100 million considered during further reduction. It also records existing stock of 650 million notes, including about 620 million remaining from the earlier contracts.

These are five documented decision stages containing several distinct figures: original contracts, later requirement, later order, revision and possible further reduction. The stock statement shows that notes from the earlier contracts existed, but none of the procurement quantities above is itself a verified count of physical production. Calling any one of them “the print run” would erase the decisions that explain why the numbers differ.

Printed, Delivered and Issued Are Separate Measurements

The Eurosystem shows how scope can change even when the denomination and year appear identical. The ECB’s production page says the French, German and Italian central banks produced 1.7 billion €50 banknotes in 2017. A separate production table records 3,300 million €50 notes for 2017 and lists fifteen national central banks commissioning production. The first figure is explicitly a three-bank subset; the second belongs to the wider programme. The ECB does not state that the two use an otherwise identical basis or reconcile them, so no percentage share is calculated here.

Some official sources do use language that answers the production question directly. A Reserve Bank of Australia study of the new five-dollar note states that Note Printing Australia had printed 172 million notes before the design was issued on 1 September 2016. That is a physical production figure tied to a denomination and issue project. It should still be recorded as production before release, not as 172 million notes circulating on the first day.

Delivery is a different event. The RBA’s 2023 annual report says Note Printing Australia delivered around 601 million banknotes under contract to Singapore, Papua New Guinea and the Philippines during 2022/23. The wording is precise but aggregated: it does not divide the total by country, denomination or design, and it does not say how many each customer later issued.

Issue is different again. When the Bank of Canada launched its commemorative ten-dollar note in June 2017, it announced that only 40 million notes were being issued. That is a prospective release quantity stated by the issuer, not a later audit of completed printing. Banxico’s CM24 statistics provide monthly counts of new banknotes put into circulation; for August 2026, the denomination rows total 69.5 million pieces, all assigned to banknote family G. The series measures release of new notes, not production and not the total stock outstanding.

Even the word issued needs context. The Swiss National Bank reports two separately labelled 2024 figures: 244.0 million notes issued at its cashier’s offices, and 41.5 million freshly printed banknotes put into circulation. The source does not describe the second figure as a subset of the first. They record differently defined activities, which is exactly why the label must travel with the number.

Why Serial Ranges Do Not Automatically Reveal a Print Run

A serial range appears to offer a simple calculation: subtract the first number from the last and add one. That gives the number of possible integers in the interval. It does not prove that every number was printed, passed quality control, was issued, or even belonged to one uninterrupted run. The serial number identifies a note; it is not a production certificate.

The Reserve Bank of New Zealand publishes official ranges for Series 7 and immediately supplies the missing caution. Not every denomination is produced every year, because production follows demand, and its quality-assurance programme may require some serial numbers not to be issued. The table is valuable evidence of prefixes and ranges. It is not permission to treat every possible number as a surviving or issued note.

A catalogue example from Mozambique makes the arithmetic visible. One stated range, 2,785,001 to 3,110,000, contains 325,000 possible numbers, while the listing records 320,000 notes — a difference of 5,000. The same chapter explains that gaps of 50,000, 10,000 and 5,000 were inserted between runs. Another interval, 3,150,001 to 3,585,000, contains exactly 435,000 numbers and matches its recorded quantity. The first mismatch is not evidence of bad arithmetic; it is evidence that a serial span and a note count answer different questions.

Prefixes create further risk. A prefix can change with denomination, year, printer, replacement system or administrative practice. Before calculating anything, establish how the issuer used the prefix and whether skipped or reserved numbers are documented. The method used to identify replacement banknotes is relevant here: a distinctive prefix may describe a special production role without revealing the size of that production.

Outstanding, Destroyed and Graded Notes Are Different Populations

Once notes enter circulation, the population changes continually. New notes are released, usable notes return to banks and go out again, worn notes are removed, and some currency is lost, exported, saved or destroyed outside official channels. An outstanding figure is therefore a stock at a specified time, not the sum of every note ever printed.

The Riksbank’s 2024 annual report gives annual-average numbers of banknotes in circulation: 262 million in 2020, 256 million in 2021, 265 million in 2022, 256 million in 2023 and 245 million in 2024, each rounded. The 245 million figure describes all Swedish banknotes in circulation on an annual-average basis. It cannot be assigned to one denomination, date or catalogue variety.

Hong Kong shows why percentages and values cannot manufacture a missing count. The Hong Kong Monetary Authority’s 2023 annual report divides banknotes by number across six denominations; the published shares add to 100.0%. A separate chart gives the value of banknotes in circulation as HK$592,585 million. Multiplying a denomination’s share by the total value would mix a share of pieces with a value stock. Without an absolute count or denomination-level values, the total number of banknotes cannot be derived.

Removal figures form another population. The Swiss National Bank reports 30.1 million damaged or recalled banknotes destroyed in 2024. An ECB data series records 219.73 million euro banknotes sorted as unfit in April 2026. Sorting to unfit is a quality decision and a precursor to removal, not by itself a direct observation of physical destruction. Neither number can reconstruct the original quantities of the affected issues.

Finally, a grading population counts submissions. PMG describes its report as a tally of notes it has certified by type and grade and expressly says the figures are not indicators of rarity or value. Ungraded survivors are absent, and repeated submissions or resubmission controls belong to the grading service’s own data rules. A population report can describe the certified market. It cannot establish how many notes survive in private hands.

What a 225-Chapter Catalogue Review Actually Found

To see how specialist catalogues handle quantities in practice, this study examined all 225 supplied country chapters of The Banknote Book: 11,075 pages and 3,729,842 extracted words. Candidate figures were filtered, grouped with their catalogue context and checked through a 300-row stratified human validation sample. The full registry confirms that the number token was recovered with its source context; the sample, not the full census, supports the semantic accuracy of the classifications.

  • 6,221 identifiable banknote counts were retained: 6,217 stated directly and four completed from clear elliptical wording.
  • 5,510 of 6,221, or 88.6%, carry no measurement label. They state a number of notes but do not say whether it is production, issue, survival or another category.
  • Only 32 counts, about 0.5% or one in 194, are explicitly labelled as physically printed.
  • The exact phrases print run and print runs occur only 15 times across 11 chapters: 11 singular occurrences in seven chapters and four plural occurrences in four different chapters.
  • 163 chapters contain at least one identifiable count; 62 contain none. The distribution is highly uneven: Scotland alone supplies 2,554 counts, or 41% of the total.

The absence of a label does not make a catalogue number wrong. It limits what can safely be claimed from it. “40,000 notes” may be exact as transcription and still leave the measurement unknown. The correct public wording is therefore the catalogue records 40,000 notes without identifying the measurement, not the print run was 40,000.

The extraction also found why automation alone is dangerous. Tens of thousands of numeric expressions came from auction advertising, graded-lot descriptions and bibliographies rather than banknote quantities. One repeated advertisement referred to millions of online bidders. Without context and exclusions, a scraper could promote an auction platform’s membership figure into a supposed print run.

These results describe one specialist catalogue reference set reviewed on one date, not every catalogue and not every banknote ever issued. A phrase not found in the reviewed chapters is reported as not found there; it is not claimed never to exist elsewhere.

How Collectors Should Record a Banknote Quantity

A reliable record preserves the path from source to claim. The following workflow works for a catalogue annotation, research note, auction description or collection database.

  1. Identify the exact object. Record country or issuing authority, denomination, series, date, signatures, prefix and catalogue variety. A national aggregate should never be attached to one variety merely because its date looks close.
  2. Copy the source’s measurement word. Preserve ordered, printed, delivered, issued, outstanding, destroyed or certified. If no label is present, write measurement not stated.
  3. Record the unit and period basis. Notes, thousands of notes, monetary value and percentages are different units. Calendar year, fiscal year, month-end, annual average and cumulative total are different periods.
  4. Record the scope and provenance. State whether the figure covers one denomination, all denominations, one printer, several customers or a whole monetary system, and whether the evidence is an issuer record, printer report, contract, catalogue or market database.
  5. Do not convert a proxy silently. Serial ranges, prefix blocks, face value, denomination shares and grading populations can support analysis but cannot become production counts without an explicit bridge supplied by the source.
  6. Keep quantity separate from rarity. Rarity depends on survival, availability and demand. A large production can become scarce after withdrawal and destruction; a small commemorative issue can remain common if collectors saved it in quantity.

Recommended form: “172 million notes physically printed for the new Australian $5 issue (Reserve Bank of Australia, before release on 1 September 2016).”

That sentence is longer than “print run: 172 million”, but it can be checked and cannot easily be mistaken for the number issued, outstanding or surviving. Precision is not decoration here. It is the difference between preserving evidence and manufacturing a statistic.

Final Thoughts

Banknote quantities are not absent from the documentary record. They are dispersed through print orders, production reports, procurement minutes, annual reports, statistical series, serial-number tables and catalogue notes. The difficulty is that each source observes a different part of the banknote life cycle, and the convenient phrase print run often hides that difference.

The best answer to “How many were printed?” may be a precise production count. It may instead be that only an order, a serial range or an outstanding stock is known. Saying which one exists — and refusing to rename it — produces a stronger catalogue than filling every blank. A banknote number becomes useful only when the measurement travels with it.

Frequently Asked Questions

What is a banknote print run?

In strict use, a banknote print run is a defined batch or quantity physically produced under a particular production arrangement. In collecting, the phrase is often used more loosely for orders, serial ranges, quantities issued or even surviving examples. The source must identify what was counted before the number can safely be called a print run.

Is a central bank print order the same as the number printed?

No. A print order states what an authority asks a printer to supply and may be a range, a ceiling or a revisable plan. Completed production is a later measurement. In the United States, the Federal Reserve publishes the order while the Bureau of Engraving and Printing publishes separate production reports.

Can serial numbers be used to calculate a banknote print run?

Only when the source confirms that the range was continuous and explains skipped, reserved, rejected or unissued numbers. Subtracting the first serial from the last and adding one gives the number of possible integers, not proof that every number became a printed or issued note.

Does a low print run make a banknote rare?

Not automatically. Rarity depends on how many notes survived, how many are available to collectors and how strong demand is. A large production can become scarce after withdrawal and destruction, while a small commemorative issue can remain readily available because many examples were saved.

Does a grading population report show how many banknotes survive?

No. It counts notes certified by that grading service, not every surviving note. Ungraded collections are absent, and submission patterns favour notes considered valuable enough to grade. PMG itself states that its population report is not an indicator of rarity or value.

Where can reliable banknote quantity figures be found?

The strongest sources are issuing authorities, official printers, annual reports, statistical databases, procurement records and contemporary release notices. Specialist catalogues can preserve important figures, but the exact entry and its wording must be read. Dealer listings and unsourced claims should not be used as primary evidence.

Selected References

Quantities below are attributed to the measurement used by each source. A number described as an order, delivery, issue, outstanding stock, unfit count or grading population is not silently converted into production.

About the Author

Audrius is the founder and curator of WorldBanknotes.eu, a specialised catalogue and shop for collectible banknotes from around the world, and a long-time collector and researcher. His work focuses on banknote identification, catalogue structure, printing, design and the historical and institutional context behind individual issues.

Follow the production chain through How Is a Banknote Created? and How Banknotes Are Printed. For the institutional boundary between manufacture and release, see How Banknotes Became Money and Banknotes That Never Became Money.

The four original scans used in this guide are catalogued individually: Brazil 1 Centavo 1967, P-183b, New Zealand 5 Dollars 2015, P-191, Sweden 20 Kronor 2015, P-69a and Hong Kong 10 Dollars 2014, P-401d.

Browse the full banknote catalogue, and record every quantity with its original measurement label before using it as evidence.