How to Test Gold at Home (and Which Tests Actually Work)

Most of the gold tests circulating online cannot answer the question people are actually asking.

Search how to test gold at home and you get a page of YouTube videos, a Quora thread and a Facebook group post, demonstrating vinegar, toothpaste, lighters, ice and salt. The demonstrations are real. The problem is what they are testing.

Almost nobody holding a gold chain wonders whether it is gold-coloured plastic. They wonder whether it is solid gold or gold plated — and that is a question most of these tests are structurally incapable of answering, because plating is real gold on the outside. Any test that only touches the surface will pass a plated piece.

Here is what each test actually proves.

The magnet test: rules things out, proves nothing

Hold a neodymium magnet to the piece. If it sticks firmly, it contains iron or steel and is certainly not solid gold.

That is the entire value of the test, and it is genuinely worth thirty seconds. What it cannot do is confirm gold, because brass, copper, lead, aluminium and tungsten are all non-magnetic — and brass is precisely what most plated jewellery is built on. A plated brass chain sails through the magnet test.

One footnote: clasps and spring rings often contain steel springs even on genuine gold pieces. Test the body of the item, not the fastening.

Verdict: useful for ruling out, useless for ruling in.

The vinegar test: fails on the one case that matters

Drop the piece in white vinegar, wait, look for fizzing or colour change. Gold does not react with acetic acid, so real gold comes out unchanged.

The flaw is simple. On a plated piece, the vinegar meets the gold plating, not the brass underneath. Intact plating does not react either. So the plated chain also comes out unchanged, and you have learned nothing about the thing you wanted to know.

It will catch bare, unplated base metal. Very little of what people are worried about is bare base metal.

Verdict: passes plated jewellery, which defeats the purpose.

Toothpaste, lighters, ice, lemon: skip all of these

These are the ones to be actively wary of, because several damage the piece.

Toothpaste is an abrasive. Rubbing it on soft gold leaves fine scratches, and whatever colour change you see tells you nothing reliable. A lighter can discolour the metal, weaken solder joints where a chain is linked, and destroy stones outright — many gemstones crack from thermal shock, and anything glued rather than set will simply let go. Ice rests on gold’s thermal conductivity, which is real, but silver and copper conduct heat similarly well, so it does not isolate gold. Lemon juice is the vinegar test with extra steps and the same flaw.

Verdict: risk your jewellery for a result you cannot trust.

The ceramic scratch test: works, but costs you

Drag the piece across unglazed ceramic. Gold leaves a gold streak; most fakes leave a dark one.

This one genuinely does cut through plating, which is why it survives in the trade. It also drags metal off your jewellery and leaves a visible mark. On a piece you intend to keep, that is a real cost for information a jeweller will give you free.

Verdict: it works, it is destructive, there are better options.

The “float test” is not a test

The version circulating — drop it in water, real gold sinks — is nonsense. Every metal sinks in water. Brass sinks. Steel sinks. A gold-plated zinc pendant sinks.

What people are garbling is the density test, which is real, which is the best home test there is, and which is almost never explained properly.

The density test, done properly

Gold is extraordinarily dense: 19.3 g/cm³ pure, around 13 for 14k, against roughly 8.5 for brass. That gap is wide enough to be decisive, and you can measure it with a kitchen scale and a glass of water.

1. Weigh the piece dry. Call that W. You need a scale reading to 0.01 g — a jeweller’s scale costs very little.
2. Suspend it fully in water without touching the sides or bottom. Thread works. Weigh again and call that S.
3. Divide: density = W ÷ (W − S).

Compare against the table:

A reading near 13 on a piece stamped 14K is consistent. A reading near 8.5 means brass with a gold coat, whatever the stamp says.

Two honest caveats. Hollow pieces break the test — hollow rope chains and puffed hearts trap air and read far too light, and they may still be solid gold. And tungsten defeats it, because tungsten is almost exactly as dense as gold, which is precisely why counterfeit bullion bars use it. For jewellery that is vanishingly rare; for a gold bar, it is the reason dealers ultrasound rather than weigh.

Read the stamp before any of this

The fastest check costs nothing. Look inside a band, on the back of a clasp, on the tag end of a chain, with a loupe or a phone camera zoomed in.

• 14K, 585, 750, 417 and similar mean solid gold at that purity — our post on what 585 means covers the fineness marks.
• GP, GEP, HGE, RGP mean plated, and the karat number in front refers only to the coating.
• 1/20 or GF mean gold filled — far more gold than plating, far less than solid.
• 925 with a gold colour means vermeil: gold over sterling silver, worth real money for its silver.

A stamp is a claim rather than a guarantee, and fakes are stamped too. But a stamp that says GP settles the question immediately, and most people never look.

Then look at the wear

Plating fails where a piece rubs. Check the inside of a ring band, the back of a clasp, the underside of chain links and any raised edge. A colour change to white, grey or pink underneath is the base metal showing through.

Solid gold cannot do this. It may discolour on the surface — the copper in the alloy oxidises, which is what our post on whether 10k gold tarnishes is about — but it cannot wear through to a different colour, because there is nothing underneath.

The answer nobody mentions: have it tested free

Every home test is an approximation of what a jeweller’s acid kit or XRF machine measures directly, in about a minute, usually free while you watch.

An acid streak test puts a mark on a testing stone rather than on your jewellery. An XRF reading touches nothing at all and returns the actual alloy composition. Either gives you the karat, not a yes-or-no — and the karat is what determines value, as our scrap page and the karat price table set out. Manhattan Gold & Silver, a refiner, publishes a reasonable rundown of what a buyer runs and why.

If you are planning to sell, you will be getting it tested anyway. The home tests are for deciding whether the trip is worth making — and for that, the stamp, the wear check and the density test are the three that earn their place. The rest are content, not chemistry.