How to check if gold is real
Most buyers should not rely on one single test. The safest approach combines hallmark verification, weight and dimension checks, source quality, and where needed, a professional assay. This guide explains what each check involves, which are reliable, and which home tests are unreliable or dangerous.
Hallmarks by country
Hallmarks are the most reliable first check for retail gold. They indicate purity and are applied by the manufacturer or assay office.
For internationally traded bars (LBMA): look for refiner stamp (PAMP, Heraeus, Perth Mint), weight, fineness (999.9), and unique serial number.
Weight and dimension checks
For known products, dimensions and weight are published by the mint or refiner. Use these checks to spot obvious fakes.
Example: 1 oz American Gold Eagle โ exact weight 33.93g (contains 31.1g pure gold, remainder alloy), diameter 32.7mm, thickness 2.87mm.
A fake coin that uses tungsten (similar density to gold) would weigh correctly but fail dimension checks โ tungsten is harder to machine to exact tolerances. Use a precision scale (0.01g resolution) and vernier calipers.
Also check:
- Edge milling pattern on coins โ each mint has a specific pattern
- Surface texture โ genuine coins have crisp, consistent detail; fakes often have softer or blurry features
Home tests โ what works and what does not
- GOOD: Magnet test โ gold is not magnetic. If a piece strongly attracts a magnet, it contains significant ferrous metal. However, tungsten and copper are also non-magnetic, so passing the magnet test does not confirm gold.
- GOOD: Acid test โ a licensed gold tester applies a small amount of nitric acid to a scratch mark on an inconspicuous area. Reaction tells you the karat. This is a reliable consumer test available from jewellery supply stores (~$20 kit). Note: damages the surface slightly.
- PARTIAL: Float/sink test โ gold is very dense (19.3 g/cm3). Most fakes are lighter and may float or sink more slowly. However, tungsten (19.25 g/cm3) has nearly identical density. Not reliable for detecting tungsten-plated fakes.
- NOT RELIABLE: Bite test โ pressing gold with your teeth is not a meaningful test. Gold is soft but so are many other metals. A thin gold plating over a soft metal base passes the bite test.
- NOT RELIABLE: Colour and shine โ plated fakes can look identical to solid gold under normal lighting.
- BEST: XRF (X-ray fluorescence) analysis โ measures elemental composition non-destructively. Dealers and assay offices use handheld XRF guns that give purity readings in seconds. This is the gold standard (no pun intended) for verification.
Red flags when buying
- Price significantly below spot โ if an offer is 10%+ below the live market price per gram, something is wrong. Genuine gold does not sell below spot except under extraordinary circumstances.
- No hallmark or serial number โ unverified gold has a major resale discount.
- Seller unwilling to wait for verification โ any legitimate seller will allow time for a professional assay.
- Unusual weight โ a bar or coin that weighs 1-2g off specification is a serious red flag.
- Private or untraceable seller โ in-person transactions with strangers, online classifieds without buyer protection, or social media sales carry high counterfeit risk.
- Unsealed assay cards that look tampered with โ the assay card should be factory-sealed; a re-sealed card suggests the coin may have been swapped.