How To Tell If Silver Is Real: 12 Measured Checks

How to tell if silver is real starts on a scale: sterling reads 10.3 to 10.4 g/cm3, carries a 925 stamp and ignores a magnet. Plated brass fails the density.

Silver Jewellery, Flatware, Coins and Bars: What Changes and What Doesn’t

The fineness standard changes with the form and the country, and the fake changes with it. Sterling at 925 parts per thousand is the norm for jewellery and flatware in the US and the UK. A German fork can be 800 or 835, a Danish brooch 830 and a French tureen 950. Each is a legitimate standard. Bullion bars and coins run 999 or 9999. Pre-1965 US circulating coins are 900 and were never marked at all.

What does not change is the physics. Every one of those alloys sits between 10.3 and 10.5 g/cm3, none is magnetic, and every one conducts heat better than anything a counterfeiter can afford to substitute except copper. Nickel silver, the alloy under most EPNS plate, is 60 percent copper, 20 percent nickel and 20 percent zinc with no silver in it, and it reads 8.4 to 8.9 g/cm3. That gap is the reason a $30 scale beats a $3,000 opinion.

Read the table by row. A piece that reads 8.6 g/cm3 has already failed whatever the stamp says, and a coin that weighs 5.67 g is clad no matter how worn the date is. The cards at the top of this page give the acceptable range for each measurement; this section explains which measurement matters for which object.

Stamps That Are Not Fineness Marks

EPNS is the stamp most often mistaken for a silver mark, and it means Electro Plated Nickel Silver: a thin skin of silver over an alloy that contains none. The Silver Society notes that manufacturers placed the four letters in separate shields precisely so the group would resemble a British hallmark set. EP, EPBM (electroplated Britannia metal), A1 and Quadruple Plate are the same family. A1 is the best grade of plate and is still plate.

The German and Austrian numbers 90, 100 and 150 on cutlery are plating weights, not purities. Mozeris Fine Antiques records that 90 means 90 grams of silver deposited across a dozen forks and a dozen spoons of hotel-grade flatware. A spoon stamped 90 is a good spoon and is not sterling. “Sterling 90” does not exist as a standard and should be read as a plate mark with a word added.

Tibetan silver, Thai silver sold without a number, alpaca, German silver, nickel silver and “silver tone” describe colour and not content. An XRF survey of items sold online as Tibetan silver found six of seven were copper, nickel and zinc alloys. Wikipedia’s entry flags lead and arsenic as possible components of the undefined alloy. None of these words is a fineness claim under 16 CFR 23.5. That section bars the unqualified word silver on a coated product. It also requires plating of reasonable durability on all significant surfaces before a product may be called silver plated.

Home Tests Ranked by What They Rule Out

Each test rules out one class of fake and leaves the rest standing, so the order matters more than any single result. Run them cheapest first and stop at the first fail.

1. Magnet. Rules out steel and iron cores in five seconds. Leaves brass, copper, nickel silver, lead and cadmium alloys standing, because none of them is magnetic either.
2. Weight in hand and on the scale. Rules out any coin or bar that misses its published weight, and any jewellery piece that feels light for its size. A sterling ring is about 1.2 times heavier than the same ring in nickel silver and 1.25 times heavier than brass.
3. Density by displacement. Rules out every base alloy: brass, copper, nickel silver, steel and zinc all read 7 to 9 g/cm3 against 10.3 to 10.5 for silver. Leaves lead-cored bars (11.34, so they read heavy) and thick plate over copper, which can land near 9.5 to 10.
4. Ice cube. Rules out steel, nickel silver and pewter, which sit under 70 W/m K and leave the cube dry. Leaves copper (401 W/m K), which melts ice nearly as fast as silver at 428.
5. Ring test. Rules out lead-filled and thickly plated coins, which clunk. Leaves a plated coin of the right weight, which can still ring.
6. Bleach or sulphur tarnish. One drop of bleach blackens real silver within a minute; the Royal Society of Chemistry notes that silver tarnishes as sulphur compounds react with the surface, and hypochlorite attacks it faster still. Silver plate blackens too, so this test cannot separate sterling from EPNS.
7. Acid in a filed notch. Rules out plate, the one fake that beats everything above. The notch exposes the core; dark red means 925, green or bubbling means copper or brass. This is the only home test that reads below the surface, and it leaves a small permanent mark.

Two fails from that list end the question. A piece that passes all seven is silver of some fineness, and only the stamp, an XRF or a fire assay will say which.

Coins: Weight First, Then the Edge

A coin has a published weight, diameter and thickness, and those three numbers are locked together by density. That makes coins the easiest silver to check and the most common silver to fake. The US Mint lists the American Eagle silver bullion coin at 1.000 troy oz (31.103 g), 1.598 in (40.60 mm) and 99.9 percent silver; Gainesville Coins adds 2.98 mm for the thickness. A coin that hits 31.1 g and 40.6 mm but measures 3.3 mm thick has a core lighter than silver and needed more volume to make the weight.

Since the mid-2021 redesign every Silver Eagle carries a reed pattern variation on the edge, announced by the Mint as an anti-counterfeiting measure. Coin World tracked the notch by year, viewed from the obverse. It sits at roughly 6 o’clock on 2021 Type 2 coins, about 7 o’clock just below the sun on 2022 coins, and 3 o’clock on 2023 coins. A 2022 coin with unbroken reeding, or a notch at 6 o’clock, was not struck at West Point.

Pre-1965 dimes, quarters and halves are 90 percent silver and were never marked, so the weight is the mark. The silver standards are 2.50 g, 6.25 g and 12.50 g, and the Morgan and Peace dollars are 26.73 g. The Mint’s current clad coins weigh 2.268 g, 5.67 g and 11.34 g in the same denominations, a 9 percent gap that a kitchen scale resolves. Stand the coin on edge: a 90 percent coin shows a uniform white edge, a clad coin shows a copper stripe between two cupronickel layers.

The 1965 to 1970 half dollar is 40 percent silver and weighs 11.50 g, close enough to the 11.34 g clad half that the scale alone will not separate them; the edge tone and the date do. Hero Bullion puts its silver content at 4.6 g.

Where To Get Your Silver Authenticated

A jeweller or pawn shop with a handheld XRF gives you a fineness reading in about ten seconds, usually free, and that reading is the right thing to ask for. XRF reads roughly 10 to 50 microns into the metal, so a heavy plate can return a silver result at the surface. On a piece you suspect is plated, ask for a second reading inside a filed notch. Ask for the number, not a yes or no.

In the UK the four assay offices (London, Birmingham, Sheffield and Edinburgh) test and strike the hallmark itself. Hallmarking is compulsory on silver over 7.78 g under the Assay Office London guide. The Birmingham Assay Office also runs the laboratory that measured the cadmium in the Sunday Times samples, and it takes private submissions. In the US there is no public assay office; a refiner’s fire assay is the destructive equivalent and is used when metal is being sold for melt.

Coin and bullion dealers carry the instruments jewellers do not. A Sigma Metalytics verifier reads resistivity through the metal and through a plastic flip, and a dealer who buys silver runs one on every bar across the counter. PCGS and NGC grade and encapsulate collectible coins, which moves the question from the coin to the holder. For silver bought at a live-stream auction, this site’s Whatnot review sets out what the platform’s buyer protection covers on precious metals. The gold guide on this site uses the same density method with different target numbers, and the silver hub and the wider jewellery category collect every related page. Where a 925 punch sits on a branded piece rather than on flatware, the Chrome Hearts guide shows what the stamp has to agree with before it means anything.

Legitly reads your photos before any of that. It checks the stamp format and placement, the hallmark layout, the colour at wear points and the edge of a coin. It flags what does not fit the standard for that kind of piece. A photo cannot weigh a spoon or read the inside of a notch, so Legitly pre-screens and never certifies. Use it to decide whether a listing deserves your time, then use the scale before your money moves.

Verdict: One Stamp Proves Nothing, Two Numbers Decide

Genuine silver passes every card on this page, so a single pass means little and a single fail means a lot. Fakes are built to pass the checks buyers run most: a 925 punch costs nothing, a non-magnetic alloy costs nothing, and a flash of silver plate survives a surface acid drop. What a fake cannot do is pass density and a filed notch together, because a brass or nickel silver core is 15 to 30 percent too light and turns the acid green.

Rank the tests by what they cost the counterfeiter to beat. A stamp is free. A magnet pass is free. A thick plate that beats the ice test and the ring test costs a little more. A correct density costs the price of silver or the price of lead, and lead only makes sense inside a bar. That is why the scale is the home test that matters for jewellery and flatware, and why a bar needs a dealer’s resistivity or ultrasound reading before the balance is paid.

Two fails end it. A 925 ring at 10.35 g/cm3 with a dark red notch is sterling. A 925 ring at 8.6 g/cm3 is not, and the acid is a formality. A Silver Eagle at 31.2 g with no edge notch, or a 1964 quarter at 5.67 g with a copper stripe, has failed twice. Walk on two fails, whatever the story, the box or the price.