GoldInvestingHow to Tell if Gold Is Real, and Why a Magnet Isn't Enough
To tell if gold is real, check its weight and size against mint specs, try a magnet, read the assay card, then have a dealer run an XRF or conductivity test.
You tell if gold is real by checking its weight and dimensions against the mint's published specs, trying a strong magnet, looking hard at the design and packaging, and then having a dealer confirm it with an X-ray fluorescence (XRF) or conductivity test. No home test settles it on its own, because the best fakes have a tungsten core that weighs almost exactly what gold does. The step that matters most comes before any test: buying from a seller whose record you've checked.
What fake gold coins and bars look like
The Deutsche Bundesbank's coin analysis center, presenting at an LBMA conference, sorted counterfeits into two groups.
The first group has real gold in it, typically between 450 and 930 parts per thousand, just less than the genuine coin. These fakes give themselves away in the minting, especially the edge lettering and the number of edge reeds (mills). The Bundesbank's examples include German 20 Mark pieces, Swiss 20 franc "Vreneli" coins, British Sovereigns and Mexican 50 peso coins.
The second group has a non-gold core, usually steel, zinc or a copper alloy, and "new versions have a tungsten core." They're normally coated in a gold layer under 0.03 mm thick. The Bundesbank says size, weight or magnetism catches them easily, "except tungsten."
Bars get faked too. An article in LBMA's Alchemist journal reports that in 2019 "more than 1,000 counterfeit gold bars, worth over $50 million, were discovered in vaults around the world." They "carried forged refinery stamps, matched expected assay and weight specifications," and some tested at 99.98 percent purity. The article describes counterfeiters using other metals with dense cores to keep the weight up while cutting the gold.
Tungsten weighs almost what gold does
It weighs almost the same as gold. A paper presented at the LBMA Assaying and Refining Conference in 2015 gives gold's density as 19.30 g/cm³ and tungsten's as 19.25 g/cm³, so a one-ounce tungsten piece can match a gold coin or bar in size and weight.
Here's what that looks like in practice. The Bundesbank measured a counterfeit U.S. $50 gold coin with a tungsten core against a genuine one:
| Measurement | Counterfeit | Genuine |
|---|---|---|
| Weight | 33.919 g | 33.947 g |
| Diameter | 32.71 mm | 32.69 mm |
| Thickness | 2.76 mm | 2.82 mm |
| Density | 15.89 g/cm³ | 17.62 g/cm³ |
| Edge reeds | 160 | 161 |
| Composition | Tungsten core, thin gold layer | 916.7 fine gold, rest copper and silver |
Weight within 0.03 grams. Diameter within 0.02 mm. A kitchen scale and a ruler wouldn't catch it. What did catch it was the density, the thickness and the conductivity reading, which the Bundesbank recorded as "disturbed."
How can you test gold at home?
Home tests can prove a coin fake. They can't prove it real. Treat them as a first screen.
1. Weight and dimensions against mint specifications
Weigh the coin on a scale that reads to at least 0.01 gram, measure the diameter with digital calipers, and compare with the issuer's figures:
| Coin | Fineness | Total weight | Diameter |
|---|---|---|---|
| American Gold Eagle (1 oz) | 91.67% gold | 33.931 g | 32.70 mm |
| American Gold Eagle (1/2 oz) | 91.67% gold | 16.966 g | 27.00 mm |
| American Gold Buffalo (1 oz) | 99.99% gold | 31.103 g | 32.70 mm |
| Canadian Gold Maple Leaf (1 oz) | 99.99% gold | 31.11 g | 30 mm |
If it's light, heavy, too wide or too thin by more than a small tolerance, be suspicious. If it matches, you still haven't proven anything, as the tungsten coin above shows. Gold coins has the full specs for each coin.
2. The magnet test
The Royal Mint says "Gold by its very nature isn't magnetic, so a genuine gold item shouldn't be drawn to a strong magnet." Get a strong neodymium magnet. If the coin sticks or drags, it isn't solid gold; the Bundesbank notes that steel-based fakes give themselves away through magnetism.
If it doesn't stick, all you've ruled out is a steel core. The Royal Mint points out that many non-gold metals aren't magnetic either, and the Bundesbank's magnetism check doesn't catch tungsten-core fakes.
3. Density (specific gravity)
Density is harder to fake than weight because it combines weight and volume. The Bundesbank uses a specific gravity balance based on Archimedes' principle: weigh the item in air, weigh it again hanging in water, and divide the weight in air by the difference. Its slides call this a simple way to catch poor fakes made of steel, brass, zinc or lead. A tungsten core can bring a fake close to the right density, though the Bundesbank's tungsten fake still came in well below the genuine coin.
4. The ping test
Gold and tungsten differ a lot in stiffness. The 2015 LBMA conference paper puts their shear moduli at 28 GPa for gold and 161 GPa for tungsten, and explains that a coin's resonant frequencies depend on its density, dimensions and elastic properties. So a genuine coin and a same-size, same-weight fake can ring differently when tapped. A ping test compares a coin's ring with the known ring of a genuine one; the lab version, resonant ultrasound spectroscopy, measures the frequencies precisely. We'd count a home ping test as one more screen. It won't prove anything by itself.
5. Design, edge and packaging
Put the coin under magnification beside a known genuine example or the mint's own photos. The Bundesbank's first-group fakes were caught by sloppy edge lettering and the wrong number of edge reeds.
Modern bullion has features built in. The Royal Canadian Mint lists three on its 2026 one-ounce Gold Maple Leaf: precision-machined radial lines, a micro-engraved maple leaf that shows the numeral "26" under magnification, and Bullion DNA, its anti-counterfeiting technology for Gold Maple Leaf coins dated 2014 and later. To verify a coin, the Mint sends owners to a registered Bullion DNA dealer.
Some minted bars come sealed in a card stating the weight, fineness and serial number. Costco's listing for the one-ounce PAMP Suisse Lady Fortuna bar says each bar is "individually controlled, registered" and sealed in CertiPAMP packaging with an official assay certificate and a digital certificate reached by QR code, and that the bar uses PAMP's VERISCAN identification system and iPhone app. Check that the serial number on the bar matches the card. If you plan to resell, leave it sealed, and use the maker's own verification tool, never a link a seller sends you.
Tests to skip
The Royal Mint says "vinegar and other household chemicals" get suggested regularly but "can produce misleading results." It lists acid testing and formal assay among the methods professionals use. Neither is a home test.
What a dealer's tests can see
Professional tests read what the metal is made of, inside and out, without damaging it.
| Test | What it measures | Strength | Limit |
|---|---|---|---|
| X-ray fluorescence (XRF) | The chemical makeup of the surface | Non-destructive; gives purity in seconds | Reads less than 0.05 mm deep, per the Bundesbank |
| Electrical conductivity | How well the metal conducts, compared with known values | Reads below the surface; can test through thin plastic, so sealed bullion stays sealed | Depth depends on frequency; results must be read against the right alloy |
| Specific gravity | Density from weight in air and in water | Catches poor fakes made of steel, brass, zinc or lead | A tungsten core can come close |
| Ultrasound | How sound travels or resonates through the piece | Can separate gold from tungsten | Specialist equipment |
The Royal Mint describes XRF as analyzing the metal's composition "without touching or damaging the item." It only reads the outer layer, though, so a thick gold shell over tungsten can read as gold.
Conductivity goes deeper. The Bundesbank's slides say the depth of a reading varies with frequency, from about 12 mm at 100 Hz to about 0.4 mm at 120 kHz, and that it can measure through plastic foil up to 0.5 mm. Pure gold reads about 44 MS/m and a 91.67% Krugerrand about 9.7 MS/m, so the tester has to be set to the right alloy. Handheld testers built this way include the Precious Metal Verifier models sold by Sigma Metalytics. Ask the dealer which test it uses, and ask to see the result.
LBMA Good Delivery and small bars
Good Delivery is the London Bullion Market Association's standard for large wholesale bars traded in London. Its rules describe gold bars of about 400 fine troy ounces with a minimum fineness of 995.0 parts per thousand, marked with a serial number, the refiner's stamp, the fineness and, for bars made from January 2019, the year and month of manufacture. LBMA says it regularly monitors the list of accredited refiners and removes those that fail its standards.
Nobody's keeping a 400-ounce bar at home, but the list still helps you. Retail bars from refiners on it come from firms LBMA monitors; Costco, for one, says its PAMP bar is made by an LBMA Good Delivery refiner. Look the refiner up on LBMA's current list. Just remember the 2019 fakes carried forged refinery stamps, so the stamp alone proves nothing.
Checking gold jewelry
Start with the marks. The FTC says to look for the karat quality mark and, near it, "the name or the U.S. registered trademark of the company that will stand behind the mark." No trademark next to the karat mark? The FTC's advice is don't buy the piece. The Royal Mint adds that a missing hallmark "does not automatically mean an item isn't genuine," since old or worn pieces can lose their marks.
Plated pieces confuse a lot of people. The FTC defines "gold filled," "gold overlay" and "rolled gold plate" as a layer of at least 10-karat gold mechanically applied to a base metal, and "gold flashed" or "gold washed" as a very thin electroplating. If another color shows through at worn edges or the clasp, the Royal Mint says there's a good chance it's plated. More in gold karats and purity.
How do you avoid fake gold in the first place?
Buy from people whose record you can check. LBMA's Alchemist article says "No single method offers complete certainty, particularly against well-executed fakes," and that's exactly why where you buy matters as much as how you test.
- Use an established dealer, and check its BBB and complaint record with the steps in how to choose a gold dealer.
- Take extra care with private sales, online marketplaces and anything priced well under what dealers charge. The CFTC says bullion is priced at spot plus a premium, so an ounce offered well below today’s spot price, today's spot price, should worry you.
- Keep receipts, assay cards and original packaging. They help when you sell.
- Have anything bought privately tested by a dealer before you pay, or make payment depend on the test.
Buying gold covers the rest of the buying process.
More on spotting fake gold
How can you tell if gold is real at home?
Weigh it to 0.01 gram, measure the diameter, compare both with the mint's published specs, and try a strong magnet. A coin that fails is fake. One that passes still needs a dealer's XRF or conductivity test.
Does real gold stick to a magnet?
No. The Royal Mint says genuine gold "shouldn't be drawn to a strong magnet." Plenty of non-gold metals, tungsten included, don't stick either, so passing proves nothing on its own.
How do you spot fake gold bars?
Check that the bar's weight, dimensions, serial number and refiner match its assay card and the refiner's published specs, then have a dealer test it. Tungsten-filled bars can match the weight, so conductivity or ultrasound testing is the stronger check.
Can a fake gold coin weigh the same as a real one?
Close enough to fool a scale. The Deutsche Bundesbank measured a tungsten-core fake U.S. $50 coin at 33.919 grams against 33.947 grams for a genuine one. Its thickness and density were off.
Does vinegar tell you if gold is real?
No. The Royal Mint says vinegar and other household chemicals "can produce misleading results."
What does an assay card prove?
It's the refiner's statement of the bar's weight, fineness and serial number. Check that card and bar match and use the refiner's own verification tool, because counterfeiters have forged refinery stamps.
Sources and fact-check12 primary sources, checked October 17, 2026
- Deutsche Bundesbank, Coin National Analysis Centre (LBMA conference, 2017): Counterfeit Gold Coins
- LBMA Assaying and Refining Conference 2015: Detection of fake gold, silver bars and coins with Resonant Ultrasound Spectroscopy
- LBMA Alchemist, Issue 117: The Silent Risk, Counterfeiting in Precious Metals
- LBMA: About Good Delivery
- LBMA: Good Delivery List Rules (January 2025)
- The Royal Mint: How to Tell if Gold is Real
- United States Mint: Bullion Coin Programs
- Royal Canadian Mint: 2026 Gold Maple Leaf 1 oz bullion coin
- Costco: 1 oz Gold Bar PAMP Suisse Lady Fortuna (New in Assay)
- CFTC: Customer Advisory, Beware of Gold and Silver Schemes Designed to Drain Your Retirement Savings
- FTC: Buying Platinum, Gold, and Silver Jewelry
- Sigma Metalytics: Precious Metal Verifier
We check every figure against the original source: the mint, the regulator, the exchange or the company itself. Spotted something out of date? Tell us.
