Three tests, in this order

Telling silver from nickel

You cannot do this by eye, and people who think they can are usually sorting on tone, which tells you nothing. A silver coin can be dark and a nickel one can be bright. What works is the date, then a magnet, then a scale, and most coins are settled by the first of those.

Test one

Read the date

This settles almost everything. Each denomination has a last silver year in the Mint’s published record, and a coin dated after it contains none. The only dates that survive this test are the ones where the metal changed during the year.

The trap is that the answer is different for the five-cent coin. It stopped being silver in 1921, forty-six years before the rest. A great many people sorting a jar apply the 1960s rule to their nickels and find nothing, which is the correct outcome arrived at for the wrong reason.

CoinSilver toNo silver from
1 centnever silveralways
5 cents19211922
10 cents19671968
25 cents19681969
50 cents19671968
1 dollar19671968
2 dollarsnever silveralways

Test two

Hold a magnet to it

Silver is not ferromagnetic, and neither is the copper it is alloyed with, so a silver coin does not respond to a magnet at all. The pure nickel that replaced it is strongly ferromagnetic and jumps. The plated steel used since 2000 is magnetic too, through the plating.

That makes the magnet the whole answer for 1967 and 1968, the two dates where a silver coin and a nickel one share a date and a weight. A fridge magnet works. A stronger one makes the response unmistakable.

The magnet says nothing about two compositions in between. The 75 per cent copper and 25 per cent nickel five-cent coin, and the bimetallic toonie whose ring and core are different metals, each sit somewhere between no response and a weak one depending on the coin and the magnet. The table below shows those as not claimed rather than guessing on your behalf.

What a magnet does, by composition
CompositionMagnet
94% steel, 1.5% nickel, 4.5% copper platingattracted
99% nickelattracted
chrome plated steelattracted
99.9% nickelattracted
94.5% steel, 3.5% copper, 2% nickel platingattracted
92% steel, 5.5% copper, 2.5% nickelattracted
94% steel, 3.8% copper, 2.2% nickel platingattracted
99% nickel (minimum)attracted
93.15% steel, 4.75% copper, 2.1% nickel platingattracted
91.5% nickel, 8.5% bronze platingattracted
multi-ply brass plated steelattracted
outer ring - multi-ply nickel plated steel, insert - multi-ply brass plated aluminum bronzeattracted
95.5% copper, 3% tin, 1.5% zincnot attracted
98% copper, 0.5% tin, 1.5% zincnot attracted
98% copper, 1.75% tin, 0.25% zincnot attracted
98.4% zinc, 1.6% copper platingnot attracted
92.5% silver, 7.5% coppernot attracted
80% silver, 20% coppernot attracted
88% copper, 12% zinc (tombac)not attracted
50% silver, 50% coppernot attracted
75% copper, 25% nickelnot claimed
outer ring 99% nickel, inner core 92% copper, 6% aluminum, 2% nickelnot claimed

Test three

Put it on a scale

A kitchen scale reading to a hundredth of a gram costs very little and settles what the magnet cannot. These are the Mint’s published standard weights for the last silver standard in each denomination and the base metal coin that replaced it.

CoinSilver versionWeighsReplacementWeighsDifference
5 cents.800 1920 - 19211.167 g99% nickel 1922 - 19424.54 g-3.37 g
10 cents.800 1920 - 19672.33 g99.9% nickel 1968 - 19772.07 g0.26 g
25 cents.500 1967 - 19685.05 g99.9% nickel 1968 - 19775.05 gnone
50 cents.800 1920 - 196711.66 g99% nickel (minimum) 1968 - 19798.1 g3.56 g
1 dollar.800 1935 - 196723.3 g99.9% nickel 1968 - 198215.62 g7.68 g

One pair the scale cannot separate. The Mint gives the 50 per cent silver 25-cent coin and the pure nickel coin that replaced it as the same 5.05 grams. For that pair, and only that pair, the magnet is the test and the scale is useless.

What not to do

  • Do not use acid. Silver testing acid leaves a mark. On a coin whose value depends on its surface, that mark costs more than the test tells you, and the magnet gives the same answer for free.
  • Do not rub or file the edge. Same reason. The edge is part of the coin, and a filed one is a damaged one.
  • Do not clean it first. Cleaning a coin to see it better is the most common and most expensive mistake in the hobby. A dark original surface is worth more than a bright cleaned one, consistently and by a lot.
  • Do not sort on colour. Silver tones dark with age and nickel stays relatively bright, which means the shiniest coins in a jar are usually the ones with no silver in them at all.

Common questions

Are Canadian silver coins magnetic?
No. Silver and the copper it is alloyed with are both non-magnetic, so a silver Canadian coin will not respond to a magnet at all. The pure nickel coins that replaced them are strongly attracted, and so is the plated steel used since 2000. That single test settles most questions about a coin from the late 1960s.
Do Canadian silver coins look different from nickel ones?
Not reliably. A worn 80% silver 25-cent coin and a worn nickel one from a few years later are the same size, the same colour to most eyes and carry the same caribou. Tone is not evidence: silver can go dark and nickel can stay bright. Use the date, then the magnet, then a scale.
What should a Canadian silver coin weigh?
The Mint publishes a standard weight for every period. An 80% silver 10-cent coin is 2.33 grams and its nickel replacement is 2.07; an 80% silver 25-cent coin is 5.83 grams and its nickel replacement is 5.05. A scale reading to a hundredth of a gram separates those pairs easily. It does not separate a 50% silver 25-cent coin from a nickel one, because the Mint gives both as 5.05 grams.
Can I test a coin with acid or by rubbing it?
You can, and you should not. Acid testing and abrasive testing both damage the surface, and on a coin whose value depends on condition that damage costs far more than the test tells you. A magnet and a scale answer the question without touching the coin.
Weights, compositions and specification periods: Royal Canadian Mint published circulation coin specifications, retrieved 2026-09-04. The magnetic behaviour of nickel, steel and silver-copper alloys is a physical property of those metals rather than a figure published by the Mint, and is stated here only where it is not arguable. Methodology.