How to Tell Real Gold From Fake: Spot the Type Before You Test

Six Things People Mean by "Fake Gold"

"Fake" is a verdict that hides the evidence. Before you test anything, name the suspect, because the suspects differ in what they're made of, what they're stamped, and whether any money is left at the end.

Two of those rows aren't fake in any legal sense. A piece honestly stamped 1/20 12K GF or 925 is exactly what it says; it becomes "fake" only when someone sells it as solid gold. That distinction matters when you're deciding whether you were cheated or merely mistaken, and whether a complaint has any footing.

The cruel case is gold-filled, because it *is* partly real gold and people reasonably expect to be paid for it. On a 1/20 12K piece the gold layer is one-twentieth of the metal weight, and that layer is half gold, so pure gold comes to about 2.5% of what the scale reads. A 20 g chain holds roughly half a gram. Our gold filled calculator runs that arithmetic for any fraction and karat. Vermeil is the sneakiest for home testers, because the silver underneath is a precious metal that behaves politely under several of the checks people trust most.

What the Stamps Are Legally Supposed to Promise

In the US, the FTC's Guides for the Jewelry, Precious Metals, and Pewter Industries (16 CFR Part 23) spell out what each plating and filling term should mean. They're guidance on what counts as deceptive rather than a stamping police force, but they're the standard honest makers mark against, which makes them the best yardstick for a dishonest mark:

• Gold electroplate (GEP): gold of at least 10K, at least 0.175 microns thick measured as fine gold.
• Heavy gold electroplate (HGE): the same 10K floor, at least 2.5 microns.
• Rolled gold plate or gold overlay (RGP): a mechanically bonded layer of 10K or better that makes up at least 1/40 of the metal's weight.
• Gold filled (GF): the same construction, at least 1/20 of the metal's weight.
• Vermeil: a sterling silver base under 10K-or-better gold at least 2.5 microns thick.
• Karat marks: the stamp should match the alloy within the National Stamping Act tolerance, three parts per thousand, or seven where the tested part contains solder.

Three practical consequences follow. First, a plating code is a confession, not a trick: GP, GEP, HGE, RGP, and GF all mean the piece is not solid gold, however good it looks. Second, the guides say a piece with a hollow center shouldn't be described as "solid," so a listing promising "solid 14K" on something built from light, puffy tubing is making a claim you can hold the seller to. Third, when a quality mark applies to only part of an item, the guides expect the item to say which part. A 14K clasp on a chain that is a slightly different color from the clasp is the classic way a stamp tells the truth about the wrong object.

For foreign marks, assay-office symbols, and the fineness numbers from 375 to 999, use the decoder in our guide to gold hallmarks.

Visible Tells, Item by Item

Different objects fail in different places. The same counterfeiter's habits look different on a chain than on a coin, so read the piece according to what it is.

Chains

Chains are where misstamps and plating live, because the stamp sits on a tag or clasp that can be swapped or bought separately. Watch for:

• A karat stamp only on the clasp, with links that don't quite match its color
• "Italy" or a maker's code with no karat or fineness number beside it (a country of origin is not a purity claim)
• A chain that feels improbably light for its width, which usually means hollow tubing: legal if disclosed, misleading if sold as solid
• Dents or crushed links, another sign of hollow construction and of less gold than the size suggests
• Pink, gray, or silvery color at the link pivots, where links grind against each other all day

Rings

Rings wear in one predictable spot: the underside of the shank, the part that rubs the neighboring finger, the steering wheel, and the gym bar. A pinkish, silvery, or gray patch there on an otherwise yellow ring is plating giving up. Look inside the band for the mark; 925 next to gold color means gold over silver, not gold. A seam or faint color shift across the bottom of the shank usually means an old resize, which is normal on genuine rings and is also why some real rings lost their stamp entirely. On cheap men's bands sold online, read the listing language before the metal: "gold tone" and "gold color" describe a color, not a material.

Coins

Bullion coins hand you an advantage jewelry never does: the issuing mint publishes the exact weight, diameter, and thickness, so a fake has to match all three at once while being the right metal. Tells include soft or mushy detail in fine areas like hair and lettering, a weight that's off the published figure by more than your scale's error, uneven edge reeding, and a color that's wrong for the alloy (22K coins look warmer than 24K ones because of their copper and silver). Coins also carry a second kind of fake: a genuine gold coin with an altered date or mint mark, made to steal a collector premium. That's a numismatic question for a dealer, not a metal question.

Bars

Bars are where the expensive fakes are. Tungsten's density is within half a percent of pure gold's, so a cored bar weighs right in the hand and on the scale. The tells live in the packaging and paperwork: a missing or mismatched assay card, a serial number on the bar that doesn't match the card, packaging that has been opened and resealed, a refiner logo that doesn't match the refiner's own published designs, and an asking price below spot. Nobody sells real gold below spot to a stranger for long.

Where It Came From Changes the Odds

Source doesn't prove anything, but it tells you which row of the table to suspect first.

Estate sales and inherited drawers

Estate pieces lean toward honest confusion rather than deliberate fraud: gold-filled lockets and watch cases from decades when GF was everyday jewelry, costume pieces stored in the same box as the real thing, and older unstamped gold that was never marked in the first place. The family's memory ("she only wore real gold") is a hypothesis, not a finding. Price is the other tell. A seller asking a small fraction of melt for a stamped 18K piece either doesn't know what they have or knows exactly what they have.

Online marketplaces

Online, the photos are the seller's and the stamp in the picture may not be on the item that arrives. Read the listing literally: "gold tone," "gold color," "18K plated," "gold filled," and "stainless steel with gold PVD coating" are all honest descriptions of not-solid-gold that skim as gold in a search result. Before paying, ask for a close-up of the stamp and a photo of the item on a scale showing grams. A seller who won't provide both has answered your question. Pay with a method that has buyer protection, and check the piece the day it arrives, inside the return window.

Travel purchases

Abroad, the usual surprise isn't fake gold but unfamiliar gold. Much of the jewelry sold across the Middle East and South Asia is 21K or 22K, stamped 875 or 916, and it looks warmer, more saturated, and softer than American 14K. That color difference is not a warning sign; it's what high karat looks like. The real risks are the street seller you can't find again and the price that undercuts the day's melt value. Buy from a shop that writes karat and weight on the receipt, and keep the receipt with the piece.

Decision Tree: From What You See to What It Probably Is

Start at the top and stop at the first line that matches. The verdicts are probabilities, not proof; the next section names the test that turns each one into an answer.

Notice what the tree never asks: how heavy it feels, whether it looks expensive, whether it marked your skin. Those impressions are exactly what counterfeiters design around, and none of them sorts one row from another.

The One Test That Settles Each Suspect

You rarely need more than one test once you know what you're looking for. Here is the deciding check for each suspect; the step-by-step procedures, safety notes, and costs live on the methods page.

• Counterfeit solid (brass or zinc alloy): density, because every gold alloy is far denser than any base metal a counterfeiter can afford, and the water-displacement measurement takes ten minutes with a kitchen scale.
• Plated steel: a strong magnet on the body of the piece, not the clasp, which the magnet section of the methods guide explains.
• Electroplate over brass or copper: acid placed inside a small filed notch, since surface acid only reads the gold skin; the acid-kit procedure covers the notch and the safety gear.
• Vermeil or plate over silver: a free XRF scan at a coin or bullion dealer, which reports the silver underneath in seconds without a mark, as described in the professional-testing section.
• Misstamped low karat: an acid kit run at the claimed karat and one step below, which brackets the true purity within a single sitting.
• Unstamped, probably genuine: the same acid bracketing to find the karat, plus density if you also want to rule out a base-metal core (both procedures here).
• Tungsten-core bar: nothing you can do at home; a dealer's ultrasonic thickness check is the deciding test, and the methods guide explains why home tests miss it.

Gold-filled is missing from the list on purpose. If the piece is stamped GF or RGP, the stamp has already answered the question, and testing it only confirms what the maker told you.

A Worked Case: The "Solid 18K" Estate-Sale Bracelet

Here's the identification approach applied to one object, with the money at the end.

A link bracelet turns up at an estate sale for $60, stamped 750 on the clasp tag, weighing 24.6 g on a kitchen scale. The stamp claims 18K, which is 75% gold. If true, that's 24.6 × 0.75 = 18.45 g of pure gold. Assume gold at $4,400 per troy ounce; at 31.103 g to the troy ounce that's $141.46 per gram, so melt value is 18.45 × 141.46 = $2,610. A $60 asking price for $2,610 of metal should trigger suspicion, not excitement.

Walk the tree. No plating code, no 925. The stamp sits only on the clasp tag, so the first question is whether the links match it, and under a loupe they do: same yellow through the link pivots, no pink bleed at the edges. Nothing about the weight says hollow. That leaves the one row that matters, "right look, bought cheap from a stranger," which is a choice between genuine 18K and counterfeit solid. The deciding test for that pair is density, so it's the only one worth running. The bracelet displaced 1.59 cm³ of water: 24.6 ÷ 1.59 = 15.5 g/cm³, right where 18K belongs and nowhere near brass at about 8.5. Verdict: genuine, with no acid and no scratch.

Now the part sellers skip. That $2,610 is melt value, not an offer. Using the site's standard payout ranges, an online refiner at 70–90% of melt lands between $1,827 and $2,349; a local coin or bullion dealer at 65–85% offers $1,696 to $2,218; a jewelry store at 50–75% says $1,305 to $1,957; and a pawn shop at 40–60% opens around $1,044 to $1,566. Same verified bracelet, a spread of more than $1,300 depending only on who you hand it to.

When the "Fake" Is Real Gold: The Misstamped Chain

The fake that costs careful buyers the most isn't brass. It's genuine gold of a lower karat than the stamp claims, because it passes every real-versus-fake check: it isn't magnetic, its density is in gold territory, and it doesn't wear through to another color. Only a purity check catches it.

Take a 20 g rope chain stamped 14K, bought from a private seller at a price based on that stamp. Keep the same assumed $4,400/oz ($141.46/g):

1. As 14K (58.33% gold): 20 × 0.5833 = 11.666 g pure, worth 11.666 × 141.46 = $1,650 in melt.
2. As 10K (41.67% gold): 20 × 0.4167 = 8.334 g pure, worth 8.334 × 141.46 = $1,179 in melt.
3. The gap: about $471 of metal that the stamp promised and the chain doesn't contain.

The tells are subtle. A misstamped chain may look slightly paler or more coppery than other 14K pieces you own, and its stamp often sits on a generic tag rather than a maker's mark. Unbranded chains and mixed lots are where this shows up, which is why an acid check at the claimed karat is worth running on any chain bought from someone you can't call back. If you're the seller, it matters too: a buyer's test will find the real karat, and the offer will be built on it.

Red Flags That Have Nothing to Do With the Metal

Sometimes the clearest evidence is the behavior around the piece rather than the piece itself. None of these proves a fake on its own, but each one should move you up the table toward the suspicious rows:

• A price below melt. Anyone selling real gold can get close to melt from a refiner in a week. A stranger offering less is pricing the item like the thing it is.
• "Solid" on something light and puffy. As above, hollow gold is legitimate but shouldn't be sold as solid. The word is doing work the weight can't.
• No testing allowed. A seller who won't let you magnet, weigh, or scan an item before money changes hands has told you something. The reverse holds when you're selling: a buyer who won't test in front of you is one to walk away from.
• Urgency. "Another buyer is coming at four" exists to stop you from doing the ten minutes of checking that would end the conversation.
• Stamp photos that don't match. On a listing, a stamp close-up with a different background, lighting, or link style from the full-item photo may belong to a different item.

The whole point of learning how to tell real gold from fake by identification rather than by brute-force testing is that most pieces confess early. The stamp, the wear points, the item type, and the circumstances narrow six possibilities to one or two, and a single well-chosen test finishes the job.

Frequently Asked Questions