Gold price used here to place value: pure gold around €125/g in early July 2026 (so 18-carat around €90/g at spot, a bit less at buy-back). The price moves every day — check it on the date you're reading.
You've spotted a pretty chain in a second-hand listing, or a vintage ring from a private seller. The price is gentle, the photos are nice, and the seller swears « it's the real thing, 18 carat, it came from my grandmother ». And you hesitate: if it really is gold by the gram, it's a good deal; if it's plated, you're paying costume-jewellery price for gold. Before you hand over your money, there are a few simple things you can do yourself — and a few « tricks » from videos that you'd do better to forget.
The problem: everyone who tests your gold… wants to buy it from you
Type « how to spot real gold » into Google. You land, over and over, on gold buyers and counters. They know perfectly well how to test a piece — it's their trade — but the day you bring them your gold, they're both judge and buyer. Their interest isn't to tell you what it's really worth, it's to buy it from you as cheaply as possible.
I have nothing to buy from you. My thing is gold by the gram. So here are the tests you can run at home, what they actually prove, and above all where they stop. Because the point isn't to become an expert in two minutes: it's to rule out the obvious fakes before you buy, and to know when a doubt deserves a real certification.
The 5 tests you can run yourself
The hallmark, your first reflex. In France, any gold piece over 3 grams sold in the trade must carry a guarantee hallmark. For 18-carat gold (750 thousandths), it's the eagle's head; you may also see « 750 » or « 18K ». Take a loupe and look for it on a chain clasp, the inside of a ring, an earring post. No hallmark on a piece that should have one is a first red flag. Be careful though: a hallmark can be imitated or transferred onto a fake — it's a strong signal, not absolute proof.
The magnet, to clear out the crude fakes. Gold isn't magnetic. Bring a powerful magnet close (not the fridge one): if the piece sticks to it, there's iron inside, so it's not solid gold. It's fast and it eliminates the silliest counterfeits. But — and this matters — passing the magnet test proves nothing: brass, copper, many plated metals aren't magnetic either. The magnet says « no » reliably, it never says « yes ».
Density, the home test that's hardest to fake. This is the most telling one, and all you need is a precise scale (to a tenth of a gram) and a glass of water. You weigh the piece dry, then weigh it submerged (or measure the displaced water) and work out its density. 18-carat gold sits around 15.5 g/cm³, pure 24-carat gold at 19.3 g/cm³ — well above most of the metals used to fake it. A plated piece or a cheap alloy gives you a clearly lower number. It's physics: hard to disguise.
The acid test, the pros' reference. The nitric acid test is the most reliable method, the one buy-back counters use. You scratch the metal lightly and apply a drop of acid calibrated for the purity being tested: no reaction = the purity checks out. Kits are sold commercially. Two big caveats: the acid marks the piece (so avoid it on something you want to keep intact), and you're handling acid at home — gloves, goggles, caution. For a second-hand purchase, it's often more sensible to leave this test to a professional.
Ceramic and sound, the supporting clues. Gently rub the piece on an unglazed ceramic tile: real gold leaves a gold streak, a fake often leaves a black one. And a solid-gold piece, when you make it chime, rings clear and long, not « plastic ». Neither of these settles anything on its own — they confirm or reawaken a doubt, to be cross-checked with the others.
The 3 myths that prove nothing (and can damage the piece)
« The magnet is enough. » Wrong. The magnet rules out fakes that contain iron, that's all. A piece plated over brass or copper will pass the magnet test without a flinch — and it's still not gold. The « it doesn't stick to the magnet, so it's fine » reflex is exactly the one dishonest sellers count on.
« The vinegar test. » You see the trick going around: a few drops of vinegar, and if it doesn't react it's gold. In reality, vinegar (acetic acid) is far too weak to make a real purity test, and it can stain or attack some alloys without proving anything. At best you learn nothing, at worst you damage the piece.
« Biting the gold. » The movie gesture. Pure gold is soft, yes, but so is lead, so is tin — a tooth mark therefore proves nothing, and most 18-carat pieces are far too hard to mark anyway. You walk away with a bite mark on a piece you haven't even bought yet. Same logic for « it shines like gold »: a plated piece shines exactly the same — that's the whole point of plating.
Home tests are for ruling out fakes, never for certifying a real one. The only « yes » that holds is from a pro who stakes their guarantee on it.
What I do, myself, before buying second-hand
My blunt take: these tests are perfect for clearing out the obvious — no hallmark, sticks to the magnet, a ridiculous density: you move on without even haggling. But none of these gestures replaces a real certification. And that's where I get cautious, because I've already been caught.
I once bought a piece on a private-seller platform, and you know what? I almost daren't get it certified, for fear of finding out I've been had. That's exactly the trap of uncertified C2C (Vinted, private Le Bon Coin): you get the gentle price, but no guarantee, and the doubt that lingers. For a few euros — a certification often costs €5 to €10 at a buy-back shop — you can clear that doubt; you just have to dare, and know where to go.
That's why OrOGramme only lists official shops with a physical premises, that certify what they sell. You pay a bit more than a private seller in a listing, but you know what you're buying, and you can compare the real price per gram instead of praying. If you want to dig in, I've written in detail on how to tell real gold from costume jewellery, on the eagle's-head hallmark and what it really guarantees, and on the pitfalls of buying gold on Vinted or Le Bon Coin.
To move faster on what matters
Once you've ruled out the obvious fakes, the real question becomes the price again: for this weight and this purity, are you paying gold at the right price per gram, or the seller's markup? That's where comparing helps — seeing what 18-carat second-hand gold is really worth elsewhere, normalised to the gram, before you say yes.
Eleven second-hand 18-carat gold shops, normalised to the gram — to place the real value of a piece before you buy it.
To wrap up
Testing gold at home is mostly about not getting caught out cheaply: hallmark, magnet, density, and you've already cleared 90% of the nasty surprises. But remember that these gestures say « no » far better than they say « yes ». For the real « yes », you need a pro who certifies — or a shop that does it for you before selling. I'm telling you what I'd check in your place; the rest, your instinct and your situation decide.
Sources for the figures cited:
- Home tests (hallmark, magnet, density, acid, sound) and their limits: naya-beauty.com, accessed 17 June 2026
- Nitric acid test as the professionals' reference method: oretmonnaies.com, accessed 17 June 2026
- Vinegar and bite myths, limits of the magnet test: veracash.com, accessed 17 June 2026
- Eagle's-head hallmark mandatory for 18-carat gold (750 ‰) in France: nerinabijoux.fr, accessed 17 June 2026
- Density of 18-carat gold (~15.2-15.9 g/cm³) and 24-carat (19.3 g/cm³): cookson-clal.com, accessed 17 June 2026