Skip to content
Back to all articles
authenticationbeginnerSource-backed · index approved

How to Tell If Gold Is Real

By Ideas All Day Ltd., Publisher13 min readPublished Updated

"Real gold" in jewelry almost always means karat gold: an alloy of gold with silver, copper, zinc, or other metals, stamped with its gold content. A gold-plated, gold-filled, or vermeil piece has real gold only on the outside. A fake has none, or hides a base-metal core under a convincing color.

You can catch many plated pieces and obvious fakes at home with a loupe, a magnet, and careful reading of the stamp. You cannot measure the karat that way. Only a test on the metal itself, such as acid on a touchstone, X-ray fluorescence (XRF), or fire assay, says how much gold is actually there. Work from the gentlest check to the strongest, and stop before anything that damages the piece.

Quick checks

Use this table as a screen. One result can mislead, but several results pointing the same way are worth acting on.

Check What you may notice What it does not prove
Karat or fineness stamp 10K, 14K, 18K, 22K, or 417, 585, 750, 916 on a clasp, shank, or tag. A stamp is easy to fake, and a clasp can be gold while the chain is not.
Plating letters GP, GEP, HGE, RGP, GF, or a fraction such as 1/20 next to the karat. These marks mean a gold layer over another metal, not solid karat gold.
Wear at edges A different color showing through at clasps, links, prongs, or the inside of a ring. Solid gold also scratches and dents. Look for a color change, not just marks.
Magnet A strong pull suggests a steel or iron core or component. No pull rules out very little. Brass, copper, silver, and aluminum ignore magnets too.
Weight and feel A large piece that feels surprisingly light deserves a closer look. Hollow gold is light, and dense metals such as tungsten can imitate gold's weight.
Skin discoloration Green or black marks can come from copper in an alloy or from a worn plating. Lower-karat gold contains copper and can mark skin as well.
Vinegar, bleach, flame, biting, ceramic scratch Skip them. They prove little, and most of them damage the surface you are trying to judge.

Read the stamp first

Find every mark before you test anything. Rings are usually stamped inside the shank. Chains and necklaces are usually marked on the clasp, on a small tag beside it, or on the end caps. Use a 10x loupe and good light, and photograph each mark straight on.

Karat marks and fineness numbers say the same thing in two systems:

Karat Fineness stamp Gold content
24K 999 Fine gold
22K 916 91.6%
18K 750 75%
14K 585 58.5%
10K 417 41.7%
9K or 9ct 375 37.5%

The fineness numbers are parts per thousand. Each one has its own page in our jewelry stamps and hallmarks guide, for example 585, 750, and 375.

In the United States, the FTC Jewelry Guides set what these words may mean. The unqualified word "gold" belongs to fine, 24 karat gold. An alloy has to carry its karat, and the guides allow a tolerance of only three parts per thousand on an article without solder, or seven parts per thousand with solder. The same rules define the plating terms you may find next to a karat mark:

  • Gold filled (GF): a layer of karat gold, at least 10 karat, bonded to a base metal and making up at least 1/20 of the weight of all the metal in the article. A mark such as "1/20 14K GF" means exactly that. See GF.
  • Rolled gold plate (RGP): the same construction with a thinner layer, at least 1/40 of the metal's weight.
  • Gold electroplate (GEP, GP): at least 10 karat gold, with a minimum thickness equivalent to 0.175 microns of fine gold. See GP.
  • Heavy gold electroplate (HGE): a minimum equivalent of 2.5 microns of fine gold.
  • Vermeil: sterling silver coated with at least 10 karat gold, at least 2.5 microns thick. See vermeil.

So "14K HGE" or "18K GP" is not 14K or 18K gold jewelry. It is a gold layer of that karat over something else. Our gold plating versus gold filled versus solid gold guide compares the three in detail.

In the United Kingdom, a hallmark is more than a stamped number. A full UK hallmark has three compulsory parts: a sponsor's mark identifying who submitted the piece, a millesimal fineness mark such as 375 or 750, and an assay office mark, such as the leopard's head of London. The Birmingham Assay Office states that hallmarking is a legal requirement in the UK for gold items weighing 1 gram and over, and that an item has to be independently examined before it can be hallmarked. A number on its own, without an assay office mark, is a maker's claim rather than a tested hallmark. Marks from other countries follow different rules, which our hallmarks by country guide explains. In India, gold sold by a BIS-registered jeweller carries a BIS hallmark: the BIS logo, a purity grade such as 22K916, and a six-digit HUID you can check in the BIS Care app. See BIS hallmark and HUID.

A stamp is still only a claim. Counterfeiters stamp "14K" on brass, and genuine old pieces are sometimes unmarked because the stamp wore away or was removed during sizing. Treat a convincing stamp as a reason to keep checking, not as a verdict.

Look for wear and plating

Plating is thin, so it wears first where the piece rubs: the edges of links, the clasp, the backs of pendants, the prongs, and the underside of a ring shank. Under a loupe, look for a change in color at those spots, such as a pink, gray, or silvery metal showing through a yellow surface. Peeling, bubbling, or flaking is a strong plating sign. Solid karat gold scratches and dents, but the color inside a scratch matches the surface.

Look at the joins as well. A chain can have a gold clasp and a plated chain, and a ring can be sized with a strip of a different metal. Check each component separately, and note any color change at a solder seam.

The magnet test: what it can rule out

Gold is not attracted to a household magnet, and neither are the other common jewelry metals. Silver and copper are diamagnetic, which means they are very weakly repelled rather than pulled. So the magnet test only works in one direction.

  • A strong pull means the piece, or a part of it, contains iron, steel, or another magnetic metal. On a chain, check whether only the clasp spring reacts. Clasps often contain a small steel spring, and that is normal. If the body of the piece pulls, it is not solid karat gold.
  • No pull tells you very little. Brass, copper, aluminum, and many other fake-gold cores are not magnetic either.

Use a strong neodymium magnet, hold the piece on a string or let it hang, and watch for movement. A faint reaction in some white gold can come from nickel in the alloy.

Weight and density

Gold is very dense, and a solid gold piece usually feels heavier than a brass or plated piece of the same size. That feel is a clue, not a test. Hollow chains and hollow hoops are made of real gold but feel light. Gold alloys are less dense than fine gold, so a 10K piece weighs noticeably less than the same shape in 22K. And tungsten, which is used in some counterfeits, is almost as dense as gold.

A water-displacement density test needs an accurate scale, a piece with no stones, and careful arithmetic, and it still struggles with small, hollow, or mixed pieces. Equipment makers that sell gold analyzers point out that legitimate jewelry is alloyed, which makes density unreliable on its own. Leave it to a professional if the answer matters.

Tests to skip

Many popular home tests are useless or harmful.

  • Vinegar: gold does not react, but neither do many other metals, and an acid can attack plating, stones, pearls, and glue. A piece that does not change in vinegar is not proven to be gold.
  • Bleach and household chemicals: GIA warns that chlorine bleach can pit or damage gold alloys. It can also harm stones and the settings that hold them.
  • Lighter or flame: heat can discolor metal, crack stones, and burn pearls, enamel, and glue. It does not measure karat.
  • Biting: tooth marks show softness, not purity, and you will damage the piece and possibly your teeth.
  • Ceramic streak test: dragging the piece across unglazed ceramic scratches it, and the streak color is hard to read on alloys and plating.
  • Nitric acid from a hardware store: concentrated acid is dangerous to handle, gives off harmful fumes, and can ruin the piece. Testing acids belong in a professional kit used with protection.

Acid and touchstone testing

Jewelers use the touchstone method to estimate karat. GIA describes it in its bench tips: the jeweler rubs the piece on a dark testing stone to leave a streak, rubs reference needles of known karat beside it, and then applies testing acids of measured strength. Nitric acid of graded strength is used for lower karats, and aqua regia, a mixture of nitric and hydrochloric acid, for higher karats. Streaks of lower purity dissolve, and the unknown streak is compared with the references that survive.

To check whether a piece is gold at all, GIA describes filing a small notch in an inconspicuous spot to expose fresh metal and applying a drop of 18K testing acid. A strongly bubbling green reaction indicates base metal. Little or no reaction suggests karat gold, and the touchstone comparison follows.

Three limits matter. The test is an estimate by comparison, not a measurement. It touches only the metal that is rubbed off, so a gold-plated or gold-filled piece can pass if only its gold surface reaches the stone, which is why jewelers file into a hidden spot first. And it uses strong acids that require ventilation, gloves, goggles, and neutralization. That is a job for a jeweler, not a kitchen table.

XRF and fire assay

X-ray fluorescence is the standard non-destructive test. An XRF analyzer excites the atoms in the area under its beam and reads the X-rays they give back to report which elements are present and in what proportion. It takes seconds and leaves no mark. The Birmingham Assay Office describes analyzing the metal content of gold with XRF before hallmarking. Because XRF reads the area it is aimed at, different parts of one piece can give different results. A clasp, a post, and a body can be different alloys.

Fire assay is the reference method. A small sample is melted and refined until only the precious metal is left, and that is weighed. It is very accurate and measures the whole sample rather than a surface, but it destroys the sample, so it is used for bullion, scrap lots, and disputes rather than for a piece you want to keep.

Many jewelers, pawnbrokers, and gold buyers have an XRF unit and will test a piece while you wait. Ask which component they tested, and ask for the reading for each part that matters.

Chains, necklaces, and rings

Some checks are specific to the type of piece.

  • Chains and necklaces: check the clasp, the tag, the end caps, and the chain itself. Look for color change at the links that rub most, usually at the back of the neck. A heavy chain that pulls on a magnet is not solid gold, but a clasp spring can pull on its own.
  • Rings: read the whole inside of the shank, including both sides of any sizing seam. Look at the prongs and the base of the setting, where plating wears through first. A ring that has been resized may show a solder line.
  • Earrings: the back or post may be marked while the body is not. A 14K post does not make the whole earring 14K.

Our ring identifier by picture and find a necklace by picture guides show which photos capture these details.

When to pay for a test

A professional test is worth it when you plan to sell, insure, buy, or divide a piece, when the stamp and the wear disagree, or when an inherited piece has no marks at all. An XRF reading from a jeweler or gold buyer is usually enough for a single piece. A formal assay is for high-value lots or disputes. Get the result per component, and keep the record with the piece.

If you are selling, ask two buyers to test the piece in front of you, and compare the karat each one reports before you compare prices. Our jewelry value guide explains how to turn the result into a value range.

What a photo can and cannot show

A clear photo can record the stamp, plating letters, wear and color change, clasp type, and construction. It cannot measure gold content. Use the jewelry identifier by photo to read and organize the marks a picture shows, and the Jewelry Identifier app to keep notes on each piece. Availability can differ by storefront. When the answer matters, follow the photo with an XRF or touchstone test.

Does real gold stick to a magnet?

No. Gold is not attracted to a household magnet. A strong pull means steel, iron, or another magnetic metal is present, so the piece is not solid gold, unless only a steel clasp spring is reacting. No pull does not prove gold, because brass, copper, and silver do not stick either.

Does the vinegar test work on gold?

No. Gold does not react with vinegar, but neither do many base metals and many coatings, so "no change" proves nothing. Acid can also damage plating, stones, pearls, and adhesives.

Is 10K gold real gold?

Yes. 10K gold is a karat gold alloy that is 41.7% gold, usually stamped 10K or 417. It is real gold jewelry, but it contains less gold than 14K or 18K, which affects its color, its wear, and its value.

Can gold-plated jewelry pass an acid test?

It can if only the gold surface is tested. The acid reacts with whatever metal is rubbed onto the stone. Jewelers file into a hidden spot first to reach the metal underneath, and they use XRF when they need a non-destructive answer.

What does 14K GF mean?

GF means gold filled. Under the FTC Jewelry Guides, a gold-filled piece has a bonded layer of at least 10 karat gold that makes up at least 1/20 of the metal's weight. "14K GF" means that layer is 14 karat. The core is another metal, so it is not solid 14K gold.

Want more jewelry knowledge?

Download our app for quizzes, care guides, and AI identification

Get Free App