Gold has been counterfeited for as long as it has been valued, which is to say, for all of recorded human history. From the ancient practice of gold-plating base metals to modern tungsten-core fakes that can fool even experienced buyers, the art of faking gold has kept pace with our ability to detect it. If you have ever picked up a ring at an estate sale, inherited a box of jewelry from a relative, or bought something from an online seller, the question has probably crossed your mind: is this real?
After 14 years working with precious metals at Tiffany & Co. and earning GIA certification, I have handled tens of thousands of gold pieces. Some of them turned out to be brilliant fakes. The good news is that real gold has physical properties that are extremely difficult to replicate, and with the right tests, you can catch most counterfeits without leaving your kitchen table.
Here are nine tests, ranked from simplest to most definitive, that will help you figure out what you are actually holding.
1. The Visual Inspection Test
Start with your eyes. Real gold jewelry made for sale in the United States, Canada, and most of Europe is required by law to carry a hallmark stamp indicating its purity. These stamps are typically found on the inside of rings, on the clasp of necklaces, or on the post of earrings.
What to Look For
Common hallmark stamps include 10K, 14K, 18K, 22K, and 24K for karat markings in the US system. European jewelry often uses a millesimal fineness system: 375 (9K), 585 (14K), 750 (18K), or 999 (24K). Some Italian gold is stamped with a star followed by a number. British gold carries an assay office hallmark that tells you the city where it was tested.
You will want a jeweler's loupe (10x magnification) for this test. They cost about $10 online and are genuinely useful to have around. Under magnification, a real hallmark will look crisp and well-defined. Fake stamps tend to be slightly blurry, uneven, or shallow.
Important Caveat
A hallmark stamp is not a guarantee of authenticity. Counterfeiters stamp fake gold all the time. Conversely, some very old and very real gold pieces carry no stamp at all, particularly pieces made before hallmarking laws existed in a given country. Think of the hallmark as a useful clue, not a verdict.
2. The Magnet Test
Gold is not magnetic. Neither is silver, platinum, or copper. If you hold a reasonably strong magnet to your jewelry and it snaps toward the magnet or clings to it, you are looking at a piece that contains iron, nickel, or cobalt. That means it is not solid gold.
How to Do It Right
Do not use a refrigerator magnet. They are too weak to be useful. Pick up a neodymium magnet (also called a rare-earth magnet) from a hardware store. These are strong enough to detect even small amounts of ferromagnetic metal inside a gold-plated piece.
Hold the magnet close to the jewelry. Real gold will show no reaction at all. If the piece slides toward the magnet, pulls slightly, or feels like it wants to stick, there is ferrous metal inside.
Limitations
This test only catches fakes that contain magnetic metals. High-quality counterfeits made from brass, copper, or tungsten will pass the magnet test with flying colors because those metals are also non-magnetic. So a piece passing this test does not confirm it is gold. It just means it is not one of the cheapest types of fake.
3. The Skin Discoloration Test
Real gold does not react with human skin. If you wear a gold ring for a few hours and it leaves a black or green mark on your finger, the piece contains metals that are oxidizing against your skin. Green marks typically indicate copper content. Black marks can come from a reaction between your skin's acids and base metals in the alloy.
There is a catch, though. Even legitimate 10K and 14K gold contains significant amounts of alloy metals (41.7% and 58.3% gold, respectively). Some people with unusually acidic skin chemistry can get faint discoloration from real gold alloys, particularly in hot and humid weather. This test works better as a red flag for clearly fake pieces than as a definitive ruling.
4. The Ceramic Plate Test (Streak Test)
Take an unglazed ceramic plate or tile. The underside of most ceramic plates or the back of a bathroom tile works well. Drag your gold piece firmly across the surface.
- Gold streak: Real gold leaves a gold-colored streak on unglazed ceramic.
- Black streak: Fake gold, or gold-plated base metal, typically leaves a black or dark gray streak.
This test is reasonably reliable, but it will leave a small scratch on your jewelry. If the piece is valuable or sentimental, think carefully before dragging it across a tile. This is a better test for costume jewelry you suspect might be real than for heirloom pieces you are trying to verify.
5. The Float Test (Density Test)
Gold is extraordinarily dense. Pure 24K gold has a density of 19.3 grams per cubic centimeter, which makes it roughly twice as dense as lead and about 2.5 times denser than steel. Even 14K gold, which is only 58.3% pure gold, has a density around 12.9 to 14.6 g/cm3 depending on the alloy metals used.
The Simple Version
Fill a glass with water and drop your piece in. Real gold sinks immediately and sits on the bottom. It does not float. It does not hover in the middle. It goes straight down, quickly, because it is so dense.
If your piece floats or sinks slowly, it is not solid gold. However, many fake pieces are dense enough to sink too, so this test mainly catches hollow or very lightweight fakes.
The Precise Version
For a more accurate density test, you will need a kitchen scale that reads in grams and a cup of water.
- Weigh the piece dry and record the weight in grams.
- Place the cup of water on the scale and zero it out (tare).
- Suspend the piece in the water using a thin thread so it is fully submerged but not touching the bottom or sides.
- Record the new weight reading. This tells you the volume of water displaced.
- Divide the dry weight by the displaced water weight to get density.
Compare your result to the known densities of gold alloys:
| Karat | Gold Purity | Typical Density (g/cm3) |
|---|---|---|
| 24K | 99.9% | 19.3 |
| 22K | 91.7% | 17.7 - 17.8 |
| 18K | 75.0% | 15.2 - 15.9 |
| 14K | 58.3% | 12.9 - 14.6 |
| 10K | 41.7% | 11.3 - 11.8 |
If your piece falls within the expected density range for its stamped karat, that is a very strong indicator of authenticity. This test is one of the most reliable you can do at home. The one material that can fool it is tungsten, which has a density of 19.25 g/cm3, almost identical to pure gold. Tungsten-core fakes are rare but they do exist, particularly in the bullion market.
6. The Nitric Acid Test
This is where home testing starts to get serious. The acid test uses nitric acid to chemically react with metals. Real gold does not react with nitric acid. Base metals do, often dramatically.
How It Works
You can buy gold testing acid kits online for $15 to $30. They typically come with bottles of acid formulated for different karat levels (10K, 14K, 18K, 22K) and a dark testing stone.
- Scratch the piece firmly against the testing stone to leave a visible deposit.
- Apply a drop of acid to the mark on the stone.
- Observe the reaction.
If the mark dissolves completely, the piece is not real gold (or is below the karat level of the acid used). If the mark remains unchanged, the piece is at least the karat level indicated by that acid. You can work your way up through the acids to narrow down the exact karat.
Safety Warning
Nitric acid is a strong corrosive chemical. Wear rubber gloves and eye protection. Work in a well-ventilated area. Do not let the acid contact your skin. Keep it away from children and pets. If you are not comfortable handling chemical reagents, skip this test and take the piece to a professional jeweler.
7. The Vinegar Test
This is a gentler alternative to the acid test, and it works well enough for catching obvious fakes. Place your piece in a small dish and pour white vinegar over it. Let it sit for 15 to 20 minutes.
Real gold will not change color or react in any visible way. Fake gold, particularly gold-plated brass or copper, may darken, develop spots, or show green discoloration where the vinegar penetrates past the plating.
Vinegar is a weak acid (acetic acid, typically around 5% concentration), so it will not damage real gold. But it also will not catch high-quality fakes the way nitric acid will. Consider this test a first-pass screening.
8. The Weight and Feel Test
This is subjective and it takes experience, but it is worth mentioning because professional jewelers rely on it constantly. Gold has a particular heft that is immediately recognizable once you have handled enough of it.
Pick up the piece and bounce it lightly in your palm. Does it feel heavier than it looks? Gold should. A 14K gold ring will feel noticeably heavier than a brass ring of the same size and shape. Gold-plated costume jewelry often feels surprisingly light compared to the real thing.
Temperature is another tell. Gold is an excellent thermal conductor. When you first pick up a gold piece, it will feel cool, but it warms to your body temperature very quickly. If a piece stays cold for a long time after you pick it up, it may not be gold.
Obviously, you need some frame of reference for this test to be useful. Handle enough gold at jewelry shops and estate sales, and you will develop a reliable gut sense for what is real and what is not.
9. The Professional XRF Test
If you want absolute certainty and the piece is valuable enough to justify the cost, take it to a jeweler who has an X-ray fluorescence (XRF) analyzer. This is a handheld or desktop device that shoots X-rays at your piece and reads the fluorescent energy that bounces back. Each element has a unique fluorescent signature, so the machine can tell you exactly what metals are present and in what percentages.
Why XRF Is the Gold Standard
XRF testing is non-destructive. It does not scratch, dissolve, or alter the piece in any way. It provides a precise readout of gold content to within a fraction of a percent. It can detect plating, identify the specific alloy metals, and flag tungsten or other unusual materials.
Most reputable jewelers will perform an XRF test for $20 to $50, sometimes free if you are a customer. Pawn shops and gold buying businesses typically have XRF machines on site. This is the test that settles all debates.
When the Tests Disagree
In practice, you will sometimes get mixed results. A piece might pass the magnet test and the float test but fail the acid test, which could indicate heavy gold plating over a non-magnetic base metal. Or a piece might have no hallmark but pass every other test, which is common with antique and international gold.
When tests disagree, weight the more definitive tests more heavily. The density test and the acid test are more reliable than the magnet test or the skin test. If those core tests point toward real gold, you probably have real gold. If you are still unsure, the XRF machine is your final answer.
Common Fakes and How They Fool People
Gold-Plated Jewelry (GP, GEP, HGE)
These pieces have a thin layer of real gold bonded to a base metal core, usually brass or steel. Stamps like GP (gold plated), GEP (gold electroplated), or HGE (heavy gold electroplate) are dead giveaways if you know to look for them. The gold layer is typically 0.5 to 2.5 microns thick, which is thin enough that acid testing will burn through it quickly to reveal the base metal underneath.
Gold-Filled Jewelry (GF, 1/20)
Gold-filled is a step above gold-plated. The gold layer is mechanically bonded and must constitute at least 5% of the total weight. A stamp like 1/20 14K GF means the piece is 1/20th (5%) 14K gold by weight. Gold-filled jewelry can pass some surface tests, but the density test and a deep acid scratch test will identify it.
Gold Vermeil
Vermeil (pronounced "ver-MAY") is a thick layer of gold plating over sterling silver. By US standards, the gold layer must be at least 2.5 microns thick and at least 10K. Vermeil is legitimate and valuable, but it is not solid gold. The acid test will catch it because the silver core reacts differently than gold alloys.
Tungsten Fakes
These are the sophisticated counterfeits. Tungsten's density is so close to gold that it fools the density test, and tungsten is non-magnetic, so it passes the magnet test too. Tungsten fakes are mostly a problem in the gold bar and bullion market rather than jewelry, but they do exist. Only XRF testing reliably catches them.
What About Vintage and Estate Gold?
If you are buying vintage or estate jewelry, you will encounter pieces that do not conform to modern hallmarking conventions. Very old gold may carry no stamp, a worn-down stamp, or a stamp from a foreign system you do not recognize. This does not mean the piece is fake.
Pieces from the Victorian era (1837-1901) were often made from high-karat gold, sometimes 18K to 22K, and may carry British assay marks or no marks at all. Art Deco pieces (1920s-1930s) are more consistently stamped. Mid-century pieces from the 1940s through 1960s can be all over the map depending on country of origin.
When evaluating vintage gold, the density test and acid test become especially important because you cannot rely on stamps alone. If you are serious about buying estate gold, learn to read karat markings and purity stamps from different countries and eras. That knowledge pays for itself many times over.
Final Thoughts
No single test is foolproof. Each one gives you a piece of the puzzle, and the more tests you combine, the clearer the picture becomes. For casual evaluation of inherited jewelry or flea market finds, the magnet test, hallmark inspection, and density test together will catch the vast majority of fakes. For pieces of significant value, spring for a professional XRF analysis. The $30 or $50 you spend on that test could save you thousands.
At D($)MVVintage, every piece we sell has been evaluated using professional-grade methods. Our GIA certification and years of hands-on training mean you never have to wonder what you are buying. But whether you are shopping with us or digging through a box at an estate sale, knowing how to test gold yourself is a skill that will serve you for life.