How to Tell If Jewelry Is Real Gold at Home

Home gold tests exist for one reason: most jewelry is not the karat its wearer thinks it is, and plated pieces can look identical to solid ones for years. Distinguishing real gold from a plated or costume piece with what you already own at home is straightforward if you understand what each test measures, where it stops, and which combinations catch the fakes a single check misses. This guide walks through the six checks worth doing at your kitchen table (stamps, magnet, water displacement, ceramic streak, skin reaction, and acid kit), then covers the plated markings that get misread, and finally when a professional test is worth the trip.

Start With the Stamp: What the Hallmark Tells You

Almost every piece of legitimate gold jewelry sold in the United States carries a stamp on an inside surface (ring shank, necklace clasp, earring post, bracelet tag). The stamp tells you the metal’s purity in one of two systems: the karat scale (10K, 14K, 18K, 22K, 24K) or the millesimal fineness scale (417, 585, 750, 916, 999). Both mean the same thing, expressed differently. The Federal Trade Commission’s Jewelry Guides at 16 CFR Part 23 set the disclosure rules U.S. sellers must follow for karat and quality marks, including tolerances and the required base for any “gold” claim.

Here is what the common marks translate to:

A stamp is a claim, not proof. Counterfeits are stamped too. But an absent, illegible, or unusual stamp is a real signal on a piece sold as gold. Look with a jeweler’s loupe (10x) or a phone macro lens on the areas where stamps usually sit. If the piece is old, worn, or was resized, the stamp may be faint but should still be visible.

Stamps you do NOT want to see if you paid for solid gold:

• GP, GEP, HGE: gold plated or heavy gold electroplate. A microscopically thin layer of gold over a base-metal core.
• GF (or “1/20 GF”, “1/10 12K GF”): gold filled. A thicker gold layer mechanically bonded to base metal, weighed as a fraction of total piece weight. The gold coating on a “1/20 14K GF” piece is 1/20th of the total weight in 14K gold.
• RGP: rolled gold plate. Similar to gold filled but a thinner cladding layer.
• EPNS or EP: electroplated nickel silver. No gold content at all.
• Vermeil (marked “925” plus a gold color): sterling silver base with a gold coating at least 2.5 microns thick. Real silver, not solid gold.

For a deeper reference on the full range of maker’s marks, country marks, and karat systems used around the world, see our jewelry hallmarks identification guide and the karat system breakdown at gold karat system explained.

The Magnet Test: Free, Fast, and Only a First Screen

Real gold is not magnetic. If a strong magnet pulls a piece of jewelry noticeably, that piece contains ferromagnetic material (iron, nickel, or steel) somewhere in its structure and is not solid gold. A refrigerator magnet is not strong enough to be diagnostic; a small neodymium (rare-earth) magnet is the useful tool here. They cost a few dollars at hardware stores.

The catch: many counterfeit gold pieces are constructed from non-magnetic base metals (brass, copper, bronze, or tungsten). Those pieces pass the magnet test cleanly while being nothing like gold. The magnet check rules out iron-based fakes, which are the crudest category, but it cannot rule out most well-made counterfeits. Treat a passing magnet result as “not obviously fake” rather than “confirmed real.”

One useful nuance: clasps, springs, and inner posts on chains and earrings often contain steel components for strength even when the visible metal is real gold. A magnet grabbing at the clasp of a chain does not mean the chain itself is fake. Test the main body of the piece, away from mechanical hardware.

The Density Test: The Most Diagnostic Check You Can Run at Home

Every metal has a specific density (mass divided by volume) that is a physical constant. Gold is dense: 19.32 g/cm³ for pure 24K, well above almost every base metal and alloy used in counterfeits. If you can weigh a piece accurately and measure its volume, you can compare the result against the published density for the karat marked on the piece. This is the single most informative test you can do without lab equipment.

You need three items: a jeweler’s scale accurate to 0.01 gram (widely available for a modest price), a small container of water, and a way to suspend the piece in the water without touching the sides. A thin thread and a chopstick over the container works.

The procedure:

1. Weigh the dry piece. Record the weight in grams. Call this W1.
2. Fill the container with water, place it on the scale, and tare the scale to zero.
3. Suspend the piece fully in the water (not touching the container). Do not let the thread rest on the scale. Record the new reading in grams. Call this W2.
4. Density in g/cm³ equals W1 divided by W2. (W2 in grams equals the volume of water displaced in cm³, because water weighs 1 g/cm³.)

Compare the calculated density against the reference table for what the piece is stamped as:

A worked example: a ring stamped 18K weighs 8.42 g dry (W1). When suspended in water on a tared scale, the water reading goes up by 0.54 g (W2). Density is 8.42 / 0.54 = 15.6 g/cm³. That lines up with the 15.5-15.6 range for 18K yellow gold. If the same ring had returned a density around 8.5 g/cm³, that would be brass or a brass-cored fake, and the 18K stamp would be false.

Two limitations. First, hollow pieces (hollow chains, puffy earrings) will give incorrect density readings because they trap air; density testing only works reliably on solid pieces. Second, tungsten has almost exactly the same density as gold, which is why sophisticated fake gold bars sometimes use tungsten cores. Tungsten fakes will pass this test and require conductivity or ultrasonic testing to detect. For jewelry, tungsten counterfeits are rare because tungsten is difficult to work into intricate shapes; the risk is mostly on plain bars and large flat pendants.

The Ceramic Streak Test: Cheap, Minor Marking, Surface Only

Rub the piece firmly across the back of an unglazed white ceramic tile (the bottom of a coffee mug works if the base is unglazed). Real gold leaves a gold-colored streak. Pyrite (“fool’s gold”) leaves a greenish-black streak. Gold-plated brass or copper often leaves a darker or more variable streak, especially once you rub through the plating layer. The test is quick, costs nothing, and gives a clear color signal.

What it does not do: it cannot distinguish 10K yellow gold from 22K yellow gold, since both leave a gold streak. It reads only the surface, so a well-plated fake with an intact gold layer will produce a gold streak too. And it leaves a small mark on the piece, so avoid it on high-value pieces where surface condition matters.

Skin Discoloration: A Symptom, Not a Test

You may have noticed green or black marks where a ring sits on a finger or a chain sits on the neck. That reaction comes from copper or silver in the alloy oxidizing against acidic skin or hand lotions. Copper produces green marks; silver produces black tarnish. Both are common in lower-karat gold (10K and 14K contain significant copper and silver) and in gold-plated pieces where the base metal is copper.

What that means: skin discoloration proves nothing definitive. Some real 10K and 14K pieces mark skin because their alloy composition happens to react with a wearer’s chemistry. Some plated pieces mark skin because the coating wore through. Some counterfeits leave no marks at all. Use skin reaction only as one input among several, not as a standalone verdict.

Acid Test Kits at Home: Useful, Destructive, Surface-Limited

Home acid test kits are the most direct chemical test available without professional equipment. A basic kit includes a set of acid solutions calibrated for different karats (10K, 14K, 18K, 22K), a small stone tablet, and a reference guide. The process: scratch the piece across the stone to leave a small metal streak, apply a drop of the acid rated one karat above what the piece is stamped as, and watch the reaction.

• Streak dissolves quickly or turns green: the metal is below the tested karat, or it is a base-metal fake.
• Streak stays gold-colored with little reaction: the metal is at or above the tested karat.
• Streak turns milky white or brown: often brass or a copper alloy underneath a gold-color coating.

Two important limitations. Acid kits read only the surface of the piece, so a plated fake with an intact gold layer will produce a passing result until you scratch deep enough to expose the base metal. You have to scratch firmly enough on the stone to get a real streak of the underlying material, not just the plating. And the acids are corrosive; wear eye protection, work over a disposable surface, and follow the kit’s safety instructions. The scratch on the piece is small but permanent.

For the deeper comparison of how home acid kits stack up against professional XRF analyzers and conductivity testers used by dealers, see our guide on XRF versus Sigma versus acid testing. If your interest is bullion (bars and coins) rather than jewelry, our companion guide is your gold real covers the tests that matter most for those formats.

Method Comparison: What Each Test Catches and Misses

The combination that catches the most is stamp + magnet + density, with an acid test as confirmation on the streak you produce. That set covers ferromagnetic fakes (magnet), most base-metal cores (density), and karat purity (acid). It still leaves the rare tungsten-cored fake and the possibility of a very well-executed plated piece that has not been scratched through.

Common Fakes and Plated Pieces: What Actually Turns Up

Most “is this real” questions on jewelry fall into one of a few categories. Knowing which category the piece belongs to guides which test to trust.

• Costume jewelry (no gold): brass, copper, or zinc alloys with a thin decorative finish. Fails magnet if any iron is present, fails density (typically 8-9 g/cm³), fails acid (reacts strongly).
• Gold plated (GP, GEP, HGE): base metal (usually brass or copper) with a thin gold layer applied by electroplating. Passes surface tests (stamp, streak, acid) until the plating wears through. Fails density unless the base metal happens to be dense.
• Gold filled (GF, RGP): thicker gold layer mechanically bonded to base metal. Passes surface tests longer than plated pieces. Still fails density since most of the piece is base metal.
• Vermeil: sterling silver base with a gold coating. Density will land near 10.5 g/cm³ (silver), well below any real gold karat. The stamp usually reads 925 or “Sterling” somewhere on the piece.
• Karat mismarking: a piece stamped 14K that is actually 10K, or 18K that is actually 14K. Density catches this because the density difference between adjacent karats is measurable. Acid catches it because the streak reaction will not match the stamped karat.
• Tungsten-cored fakes: tungsten’s density matches gold’s, so density testing passes. These are rare in jewelry (tungsten is hard to shape into intricate designs) and more common in bar-shaped counterfeits. Detection requires conductivity or ultrasonic testing.

If any test in your home battery fails, the piece is not what it is marked as. If all your home tests pass and the piece has real financial or sentimental value, a professional verification is still the right next step.

When to Get a Professional Test

A jeweler or a coin shop can run tests that are outside the practical reach of a home setup. XRF (X-ray fluorescence) analyzers give an elemental breakdown of the surface in seconds and confirm karat without any marking on the piece. Conductivity testers (the Sigma Metalytics is common) measure through the bulk of the piece and catch plated fakes that fool surface tests. Fire assay, the destructive method used by refineries, is the definitive purity measurement but only makes sense if you are selling the piece for scrap.

Situations that warrant a professional test:

• Any piece where home tests give inconsistent results (magnet passes, density fails, or vice versa).
• Estate jewelry with no reliable provenance and any real weight of gold at stake.
• A piece you plan to sell, where an XRF printout from a dealer gives buyers confidence and can lift your offer.
• A piece you plan to insure or use as collateral, where an appraisal document is required.
• Hollow pieces, plated pieces with cracked coatings, or pieces you suspect are tungsten-cored.

Many local jewelers, pawn shops, and coin dealers will XRF or acid-test a piece for free or a small fee, especially if there is a chance you will sell to them. You can find verified shops by category in the Goldiew jewelry buyer directory, coin dealer directory, or pawn shop directory. If you own a shop and are not listed yet, adding your business is free.

Selling Gold Jewelry: Get an Independent Read Before You Walk Into a Shop

Home testing tells you whether a piece is real. It does not tell you what a buyer will pay for it. Two things move that number: the current spot price of gold on a given day, and the buyer’s payout ratio (a percentage of the melt value, usually between 60 and 85 percent for retail pieces). Testing before you sell helps in two ways: you know what you own, and you can describe it accurately when you request offers, which produces tighter and higher bids.