How To Tell If a Gold Chain Is Real: 9 Measured Checks

How to tell if a gold chain is real starts with grams per inch: a solid 14K rope runs 0.75 to 0.90 g per inch, and a hollow one runs about 0.14.

A chain is the one gold item where weight settles most of the question. A ring is too small for a kitchen scale to resolve well. A coin carries a published mint specification. A chain has a length, a width and a style, and those three fix its mass inside a narrow band. A counterfeiter can copy a stamp, a clasp shape and a mirror polish for pennies. Nobody can make brass or stainless steel weigh what gold weighs at the same size. The nine cards on this page run in that order: arithmetic first, then the loupe, then the acid.

Run the four quick checks with the chain in your hand, or on a video call with the seller holding a scale. Weigh it, measure it, divide. A chain reading 0.2 g per inch while claiming solid 14K has already failed, and the magnet can stay in the drawer.

Gold Chain Styles: What Changes and What Doesn’t

Style changes the expected weight per inch and almost nothing else about the method. Rope, herringbone, Cuban, box and figaro all carry their marks in the same two places, all use a steel-sprung clasp, and all answer the same density test. What moves between them is the target number, because weight tracks the cross-sectional area of metal rather than the width quoted in the listing.

Two rows there deserve a second look. The 3mm and 4mm rope figures come from a chain weight guide that puts a 20 inch solid 14K rope at 15 to 18 g in 3mm and at 24 to 29 g in 4mm. One extra millimetre of width adds about 60 percent to the mass. The 6mm figure comes from a retailer’s own weight guide listing a 22 inch solid 14K rope at 40 to 60 g, with the caveat that link tightness varies by maker. Neither source is an assay office. Treat both as working brackets rather than published standards, which is why that card is marked Medium confidence.

Karat moves two numbers and leaves the method untouched. 10K is 41.7 percent fine gold and reads 11.5 to 13.0 g/cm3. 14K is 58.3 percent and reads 13.0 to 14.3. 18K is 75.0 percent and reads 15.0 to 16.0. A chain stamped 18K that weighs like 14K is either mis-stamped or coated, and the water test tells you which.

Hollow construction is the variant built to break the arithmetic. The wider gold guide on this site runs density across coins, bars and jewellery. On chains the hollow question comes first, because trapped air voids a water test outright and strips five sixths of the metal while the surface still looks and stamps like solid gold.

Grams Per Inch: The Number That Sorts Chains First

Divide the weight by the length before you touch anything else. A listed 14K hollow rope, 2.8mm wide and 16 inches long, weighs 2.3 g, which is 0.14 g per inch. A solid 14K 5mm herringbone at 20 inches weighs 20.5 g, or 1.03 g per inch. Those two chains photograph alike, carry the same karat, and hold metal in a ratio of about one to seven.

The procedure takes two minutes:

1. Weigh the chain on a scale that reads to 0.01 g, clasp included.
2. Lay it flat and measure end to end with the clasp closed, in inches.
3. Divide grams by inches and write the figure down.
4. Compare it with the row for that style and width. Anything 20 to 30 percent under the bracket is a question for the seller, not yet a verdict.
5. Ask in writing whether the chain is hollow. A hollow chain sold as hollow is a fair product at a fair price.

Filled and hollow goods are not a theoretical worry. Canada’s Competition Bureau closed a nationwide jewellery inspection in April 2016 and reported in December that year that roughly 5,000 pairs of earrings had been marked and sold as solid 10 and 14 karat gold while being plastic-filled and gold-covered. Five products across Canadian retail failed their own quality markings. The distributor corrected the marks, posted notices and added information cards to boxed items. Senior Deputy Commissioner Matthew Boswell put the problem plainly: buyers lack the tools to separate filled jewellery from solid, and the regulator has them.

The seller’s answers are the cheapest evidence available. Ask for weight in grams, length in inches, width in millimetres, karat, hollow or solid, and whether the clasp is the same karat as the chain. A seller with the piece in hand can answer all six in one message. Refusal, vagueness or a switch to talking about shine is the signal, and it arrives before you have spent anything.

Where the Stamp Sits on a Chain

A chain keeps its marks in two places, and both are on the fastening rather than the links. The first is the clasp tongue, the small flat face you press to open a lobster claw. The second is a thin tag soldered to the end link beside the clasp, which is where a maker puts anything that will not fit on the tongue. Links themselves are too thin and too curved to strike, so a chain with a karat mark punched into a link is unusual and worth a longer look.

American marks read in karats. European and most Italian marks read in thousandths. 375 is 9K, 417 is 10K, 585 is 14K, 750 is 18K, 916 is 22K and 999 is fine gold. Italian chain very often adds the word Italy plus a two or three letter maker code on the same tag. The 1stdibs herringbone listing carries 14kt, Italy and AFR together. That trio is a useful pattern: a chain claiming Italian manufacture with no maker code at all is claiming half a mark.

The law behind the mark sets a tolerance rather than a promise. The National Gold and Silver Stamping Act allows the assayed metal to fall 3 parts per thousand short of the stamped fineness, widening to 7 parts once solder joints are counted. Chains are solder-heavy by construction, which is exactly why the wider allowance exists. A chain stamped 585 may therefore assay at 578 and remain lawful, and a chain assaying at 400 is not a tolerance case.

Plated and filled goods have their own wording rules, and those rules are where sellers cut corners. 15 U.S.C. 297 requires a quality mark on a plated article to be accompanied by words showing it is rolled gold plate, gold plate, gold electroplate or gold filled. The same section requires the maker or dealer applying that mark to apply a registered trademark alongside it. That trademark goes on by the same method, in type at least as large, and sits as close to the quality mark as possible. Where an article has several unlike parts, each part needs its own mark of the same pattern and size. So a clasp stamped 14K on a chain with no mark anywhere is not a minor omission. It is the pattern a plated chain leaves when the coating term has been polished off.

Reading the punch itself separates a struck mark from a cast one. Struck and laser-cut marks have straight walls, a flat floor and crisp corners under a 10x loupe. Cast marks were formed in the mould with the metal, so they show rounded shoulders, uneven depth and sometimes a faint seam crossing them. The Cartier serial number check applies the same reading to signed jewellery, where the code format matters as much as the punch quality.

Why the Magnet Test Misleads on Chains

The magnet test fails twice on chains, and in both directions. Most buyers know gold is not magnetic and expect a clean chain to ignore a magnet completely. On a finished chain that expectation is wrong. Halstead, a supplier of findings to working jewellers, states in its clasp guide that a lobster claw contains a steel spring inside its metal casing, because a precious metal spring would lose tensile strength with repeated use. The spring is steel on a solid 14K clasp for the same reason it is steel on a sterling one. So a magnet tugging at the clasp of a genuine chain is the mechanism working as designed, and a buyer who walks away at that point has walked away from real gold.

The second failure runs the other way. Type 316 stainless steel is the metal most gold-plated fashion chain is built from, and material datasheets put its density at 8.00 g/cm3. A technical reference on the same alloy notes that type 316 is not magnetic and turns weakly magnetic only after cold working. A plated 316 chain therefore sits still under a magnet exactly as gold does, while holding no gold worth weighing. Silence is not a pass.

Run the magnet as a two-part reading instead. Touch it to the clasp casing, then slide it along 50 mm of links. A tug at the clasp and stillness along the links is the expected result on a real chain. Movement in the links means a steel or nickel core and ends the check immediately. Stillness everywhere sends you back to the scale, because density is the test that separates 8.00 from 13.4. How to tell if silver is real runs the same displacement arithmetic against a different target band, and the method transfers without change.

Where To Get Your Gold Chain Authenticated

A jeweller with a handheld XRF analyser is the fastest route, and on a chain it is the weakest one. XRF reads roughly the top 15 to 20 microns of surface, which is deeper than a 0.175 micron electroplate but well inside a heavy plate. On a chain the reading is also awkward: the beam sees a curved, gappy surface rather than a flat face. Ask for the karat number rather than a yes, and ask whether a density check or a filed notch was run beside it. Two methods agreeing is the standard on a chain, not one machine.

A pawn shop or gold buyer tests to protect its own money, which makes it useful and blunt. Most weigh the chain, file a notch at the clasp and drop karat acids, then quote a price against melt. That quote is itself a reading. An offer near the melt figure for the stamped karat means their tests came back consistent with the stamp. An offer at scrap-brass money means it did not, whatever the tongue says.

A UK assay office is the only authenticating body in this list that issues a legal mark. The four offices in London, Birmingham, Sheffield and Edinburgh test by fire assay and XRF and strike the compulsory punches themselves, and the London hallmarking guide sets the exemption weight for gold at 1.0 gram. A chain heavier than that, sold in Britain without three marks, is already outside the Hallmarking Act before anyone measures the metal. For a signed piece where the marks matter more than the metal, fake Van Cleef shows what a documented weight band adds to a hallmark.

Legitly sits ahead of all of them. It reads your photographs for the stamp layout on the clasp and tag, the joint pattern down the links, and the wear at the nape. It then flags whatever does not fit the karat being claimed. No photograph can weigh a chain or hang it in water, so the app pre-screens a listing and issues nothing resembling a certificate. Use it to decide whether a chain deserves a deposit, then weigh it before the balance moves. The gold hub and the jewellery category collect the related guides on this site.

Verdict: Weigh It Before You Trust the Stamp

One failure decides a chain, and it is almost always the same one. A stamp is free to copy, a clasp shape is free to copy, and a polish costs pennies. Metal content costs the price of gold. That asymmetry means the grams-per-inch figure and the density reading carry nearly all the weight of the decision, and the loupe work behind them mostly explains a result the scale already gave.

Rank the checks by what a counterfeiter has to spend. Nothing buys the 14K punch. Nothing buys a non-magnetic chain, since plated 316 stainless is non-magnetic by nature. A few pennies buy a plate that survives a surface acid drop. Only gold buys 13.4 g/cm3 and 0.83 g per inch at once, and there is no cheap substitute at that density for jewellery-scale work.

So the rule is short. A solid 14K claim with 0.15 g per inch is a hollow chain at best. A 14K claim with a density of 8.0 is plated stainless steel. Either result ends it on its own, without the acid, without the loupe and without the seller’s explanation. Where the numbers do fit, spend two minutes on the joints and the clasp wear. A correct weight with open jump rings usually means a real chain repaired with something else.