How to Tell if Gold Is Real: At-Home and Pro Tests

No single test proves gold is real. A stamp can be faked, a magnet only rules things out, and tungsten weighs almost exactly what gold does. The reliable method is to stack tests, starting with the free ones and finishing with a test that actually reads the metal.

This guide covers each test, what it proves, what it can't tell you, and the common fakes it catches or misses.

The quick answer: a 3-step routine

1. Free checks first (2 minutes). Read the stamp with a magnifier, look for worn spots showing a different color underneath, and try a strong magnet.
2. A cheap check at home (10 minutes). For anything heavier than a few grams, do a density test with a kitchen scale and a cup of water.
3. Confirm with a real metal test. Use an acid test, an electronic gold tester or an XRF scan. Many coin shops and gold buyers will test for free, because they'd have to test it anyway before buying.

If an item fails any step, stop there. It's either not gold or a lower karat than claimed.

Comparison table: every test at a glance

1. Read the stamp (and know what it really means)

Look inside ring bands, on clasps and near the ends of chains. A jeweler's loupe or your phone's zoom helps.

What a stamp proves: if it says GP, HGE, GF or RGP, you already have your answer. It's not solid gold.

What it doesn't prove: a "14K" stamp costs a counterfeiter nothing. Under federal law, a quality mark like 14K is supposed to sit next to a maker's trademark, so a lone karat stamp with no maker's mark is a mild yellow flag. Old, antique or foreign pieces often have no stamp and are still real.

2. Look for wear and color clues

Plating wears off where jewelry rubs: clasp edges, the bottom of ring bands, chain links and bracelet edges. If you see silver, pink-copper or gray metal peeking through, it's plated.

Other tells:

• Green or black skin marks. These usually come from copper or nickel in cheap alloys. Low-karat real gold can occasionally discolor skin too, so treat this as a hint, not proof.
• Too bright and too light. A chunky chain that feels weightless is suspicious, because gold is heavy.
• Flaking or bubbling on the surface is plating.

3. The magnet test

Gold isn't magnetic. Hold a strong neodymium magnet (hardware stores sell them) next to the piece. If it jumps or sticks, it has iron or steel in it.

The catch: passing proves very little. Copper, brass, lead, aluminum, silver and tungsten are all non-magnetic, so most good fakes pass. Clasps often contain a small steel spring, so test the chain, not the clasp.

For bars, try the slide test: tilt the magnet and slide it down the bar. Real gold makes it slide slowly because of a braking effect from electrical currents in the metal. It's a fun trick but not definitive.

4. The ceramic streak test

Drag the item across the back of an unglazed ceramic tile. Real gold leaves a gold or yellow streak. Many base metals leave a black or gray one.

It's cheap, but it scratches your piece and it can't tell 10K from 18K. Gold plating can also leave a gold-colored streak if you press lightly. Skip it on anything you care about.

5. The density test (best free-ish home test)

Gold is very dense, about 19.3 grams per cubic centimeter for pure gold. Most cheap fakes are far lighter: brass and copper are around 8.5–9, silver about 10.5 and lead about 11.3.

How to do it:

1. Weigh the item dry on a scale that reads to 0.01 g. Write it down.
2. Put a cup of water on the scale and tare it to zero.
3. Hang the item from a thread so it's fully underwater without touching the bottom or sides. The reading is the weight of the water it displaces.
4. Divide: dry weight ÷ displaced-water reading = density.

Example: a ring weighs 5.00 g and displaces 0.37 g of water. 5.00 ÷ 0.37 = 13.5, which sits right in the 14K range.

Ranges overlap because alloys vary, and white gold runs differently from yellow. On items under about 3 grams, tiny scale errors swing the result a lot. Gemstones and hollow pieces also throw it off.

The big blind spot: tungsten's density is about 19.25, nearly identical to gold. That's exactly why counterfeiters use it.

6. Coins: check the specs and listen for the ping

Bullion coins are made to exact specs, so measure them with digital calipers and weigh them:

A fake made of the wrong metal can't match weight, diameter and thickness all at once. Either it's too light, or it's thicker or wider to make up the weight. Collectors use a Fisch gauge to check all three in one step.

The ping test is simple: balance the coin on a fingertip and tap it with another coin. Real gold rings with a clear tone, while many fakes give a dull clink. Phone apps can analyze the sound, but they work best when you compare against a coin you know is genuine.

For current values of these coins, see our gold coin and bar values page.

7. The acid test

This is the classic jeweler test, and kits run about $15–$30.

1. Rub the item firmly on the black touchstone in the kit to leave a streak.
2. Drip the acid for the karat you're testing (10K, 14K, 18K) onto the streak.
3. If the streak stays, the gold is at least that karat. If it dissolves, it's lower or not gold.

To catch plating, file a tiny notch in a hidden spot first so the acid reaches the metal underneath. Test from the lowest acid up: if the streak survives 14K acid but dissolves in 18K acid, you've got 14K.

Cautions: nitric acid burns skin and eats clothes, so wear gloves and glasses and work near water. Filing leaves a small mark, so don't do it to antiques or coins.

8. Electronic gold testers

Handheld testers ($150 to $500 and up) touch a probe and a small gel dot to the piece and read the karat based on how the metal conducts. They're faster and cleaner than acid and accurate on solid gold.

They share the same weakness as an un-notched acid test: they read the surface. Heavy plating over a base metal can confuse cheaper models. They also can't see inside a bar.

9. XRF: the pro-level scan

An X-ray fluorescence analyzer shoots X-rays at the surface and reports the exact makeup, such as "58.4% gold, 25% silver, 16% copper." It's fast, it leaves no mark, and serious gold buyers and refiners use it. The machines cost $15,000 or more, but many coin shops will scan your piece free or for a few dollars.

Limit: XRF reads only the outer fraction of a millimeter. Thick plating, gold filled pieces, and especially tungsten-cored bars can come back as "gold" on the surface. That's why bullion needs one more test.

10. Catching tungsten-filled bars and fake bullion

The nastiest fakes are gold bars that have been drilled and filled with tungsten rods, then capped, and tungsten slugs with a thick gold coating. They pass the magnet test, the density test, the weight-and-size test, acid and often XRF.

What catches them:

• Conductivity and resonance testers such as Sigma-style metal verifiers ($800–$1,500+). These read deeper into the metal and can often test coins right through their plastic capsule.
• Ultrasonic thickness testing, which spots a foreign core inside a bar.
• Assay or drilling by a refiner, which settles it but damages the bar.

The practical defense is buying from established dealers, and keeping assay cards and receipts. A big discount below spot is the biggest red flag in bullion. Recognized coins and bars normally sell back near spot (often 97–100%+ of spot), so nobody honest sells real gold for much less. Our guide on where to buy gold bars covers safe sources.

Common fakes and what exposes them

Real gold confirmed? Here's what it's worth

Once you know the karat, value is simple math: weight × price per gram for that karat. For example, 10 grams of 14K is worth about 10 × $73.52 in melt value at today's price. Plug yours into our scrap gold calculator or check the live 14K gold price per gram.

Gold filled has some value too. A 20-gram bracelet stamped 1/20 14K GF contains about 1 gram of 14K, roughly $73.52 of melt before any buyer's cut. Most buyers won't pay much for it.

Keep in mind that melt value isn't what you'll be paid. Large mail-in refiners typically pay 85–95% of melt, local gold buyers and coin shops 75–90%, jewelers 70–85%, pawn shops 60–80%, and gold parties and mall kiosks often only 30–60%. Our guide on how to sell gold walks through getting quotes.

The bottom line

Use the free checks (stamp, wear and magnet) to weed out obvious fakes. Use a density test for a solid at-home read on heavier pieces. Then confirm with acid, an electronic tester or XRF before you buy or sell. For bullion, the main threat is tungsten, so buy from reputable dealers and ask for a conductivity test on anything that seems too cheap. If a deal looks too good, assume it's fake until the metal itself proves otherwise.