Methodology

The sources

There are two, both published by the Royal Canadian Mint, and there is deliberately no third. A reference site assembled from several overlapping hobbyist compilations inherits every disagreement between them and can never say which one it believes.

• Canadian circulation coins: technical specifications and mintages. Royal Canadian Mint, at https://www.mint.ca/en/discover/canadian-circulation. Retrieved 2026-09-04. Seven pages, one per denomination, each carrying a technical specification list, a reverse design list and a reported mintage table. This supplies 47 specification periods, 684 denomination-year mintage rows and 62 design entries.
• Faces of the Monarch. Royal Canadian Mint, at https://www.mint.ca/en/discover/faces-of-the-monarch. Retrieved 2026-09-04. Supplies 8 royal effigy periods and the inscription notes that go with them.

Live silver and gold prices come from a market feed, and the exchange rate comes from the Bank of Canada. Those are covered separately below, because a price is not the same kind of fact as a specification.

What is covered

Every year, denomination and figure on this site comes out of the tables above. The shape of the dataset is:

• 7 denominations: 1 cent, 5 cents, 10 cents, 25 cents, 50 cents, 1 dollar, 2 dollars.
• 684 denomination-year rows, each with a page. A page exists because a row exists.
• 684 of those 684 rows carry a mintage figure, which is all of them. There is no denomination-year on this site with a blank mintage, because the Mint’s tables have no blanks within the years they cover.
• 117 calendar years, 1908 to 2024, each with a cross-denomination page.
• 231 silver-bearing denomination-years.

The last year covered differs by denomination, and follows the Mint’s table rather than the calendar. The 1-cent coin ends at 2012 because it was withdrawn from circulation. The 50-cent coin ends at 2016 because that is the last row the Mint publishes for it, not because it stopped being struck.

What is derived

Two things on this site are calculated rather than copied, and both are calculated the same way on every page.

Silver fineness is parsed out of the Mint’s own composition string during the data build. When the Mint writes “80% silver, 20% copper”, the fineness stored beside it is read from those characters. The two therefore cannot drift apart, because they are the same text.

Actual silver weight is the standard weight in grams, multiplied by that fineness, divided by 31.1034768 grams. A troy ounce is the only constant on this site that is not from the Mint, and it is a definition rather than a measurement.

Deriving rather than transcribing is the single most important decision in how this site is built. A transcribed constant is a constant that can be wrong, and it can be wrong silently, on hundreds of pages, for years. A derived one is wrong only if the source is wrong, and the source is named on every page that uses it.

One consequence worth stating: derived figures do not always land on the round numbers the trade quotes. Four 80 per cent 25-cent coins work out at 0.5998 troy ounces of silver rather than the 0.6000 a dealer will say. Both are shown where it matters, and the derived figure is never rounded to match the convention.

Named gaps

These are the places where the source does not answer the question, listed because a reference site that hides its gaps is not one.

Nothing before 1908. Canadian coins were struck from 1858, in England, until the Ottawa mint opened. The Mint publishes no specification, weight, fineness or mintage for any of those years, and neither does this site. Its design notes do reach back to 1858, and those are reproduced. The full explanation.

No 50 per cent silver period for the 10-cent coin. The Mint’s specification table for the 25-cent coin lists a period labelled “1967 - 1968” at 50 per cent silver. Its table for the 10-cent coin has no equivalent: it runs 80 per cent silver to 1967 and then 99.9 per cent nickel from 1968. We do not know why, and we do not fill it in. What that means in practice.

No mint marks before 2006. The Mint documents one mark on its circulation coinage, introduced in 2006. It publishes nothing about marks on earlier coins, so neither does this site. Mint marks.

Mintage is a count of coins made. The Mint has no record of how many coins were later returned and melted, so a mintage figure is an upper bound on survival rather than a measure of it. This matters most for the silver denominations of the 1960s, which were withdrawn in quantity once the metal was worth more than the face value.

A few mintage cells are not a single number. The Mint reports 2017 as “Classic: 0, Canada 150: 20,000,000” in several denominations, because two designs ran that year. Those cells are stored as the sum of the figures in them, and the original string is kept and shown on the page so a reader can see what the Mint actually wrote.

Overlapping specification periods are left overlapping. The Mint’s own tables cover some years twice, because the metal changed during the year. Those years show two compositions here rather than one. That is not an error in the data and it is not tidied, because tidying it would be the error: a 1968 25-cent coin genuinely can be either metal.

What is deliberately excluded

Retail and catalogue values. Canadian coin values are proprietary to the Charlton catalogue and to the grading services. This site has no licence to them and does not generate substitutes. Why, at length.

Grades and varieties. Neither is in the Mint’s published record. Both change what a coin is worth by large multiples, which is exactly why guessing at them would be harmful rather than merely incomplete.

Numismatic and bullion issues. This site covers circulation coinage only, because that is what the Mint’s circulation tables cover. Collector sets, proof issues and bullion coins are a different product line with a different record.

Magnetic behaviour where it is arguable. The site states that a magnet picks up pure nickel and plated steel and does not pick up silver-copper alloys, because those cases are not in dispute and they answer the 1968 question. It says nothing about 75 per cent copper cupronickel or the bimetallic two-dollar coin, where the honest answer depends on the coin and the magnet.

Prices and exchange rates

Silver and gold spot prices are fetched live from a market feed, with a second feed as a fallback and sanity bounds on both. If a price arrives outside those bounds, or if neither feed answers, the site renders no figure at all rather than a wrong one. The published weights are unaffected by any of that, because they do not depend on a price.

Conversion to Canadian dollars uses the Bank of Canada’s published daily average USD/CAD rate, series FXUSDCAD, retrieved from its Valet API. It is a commercial-free official rate a reader can look up. It is also a daily rate rather than an intraday one, so a Canadian melt figure on this site is a live silver price at the Bank’s last published exchange rate. Every page that shows one states the rate and its observation date.

Prices on this site are the value of metal. They are not offers, not bids, and not adjusted for any dealer spread.

How the build works

The Mint’s eight pages are downloaded once into a local mirror. A parser reads that mirror and writes a single JSON file, which is committed to the site’s repository and is the only thing the site reads at build time. Nothing is fetched from the Mint during a build.

That arrangement exists for two reasons. A build that scraped a live website would be a build that could quietly change what the site says, or fail entirely, because someone else redesigned a page. And a committed dataset can be checked: the parser ships with a set of ground-truth assertions covering the transition years, the low mintages, and the derived silver weights against the figures the bullion trade uses. Those run before the data is trusted.

One example of why that check exists. The first version of the mintage parser read every run of digits in a cell as a figure, which turned the Mint’s “Classic: 0, Canada 150: 20,000,000” into a mintage of 20,000,150 on five pages. It looked entirely plausible. It was caught by an assertion rather than by reading the page.