How to Spot Fake Metals: Home Tests for 316L, 925 and 14K

Cuff & Steel

316L steel - $49

Vintage

Steel cuff - $39

Eros

316L cuff - $59

The fastest home check for jewelry metals is a magnet: 316L stainless steel, sterling silver and solid gold are all effectively non-magnetic, so a strong pull indicates a cheap ferritic steel or a steel core beneath plating. The magnet has one important blind spot - brass is also non-magnetic, so it cannot distinguish solid gold from gold-plated brass. That gap is closed by weight, because gold is roughly two and a half times denser than stainless steel and far heavier than its size suggests.

Most guides to spotting fake metal list tests without saying what each one can and cannot prove, which is how people end up confident about the wrong conclusion. A magnet is genuinely useful and genuinely limited, and knowing where the limit sits matters more than knowing the test exists.

So this page gives five checks in the order you should run them, states the blind spot in each, and settles two myths that circulate widely enough to have cost people real money.

The Quick Answer

The Five Tests, in Order

1. The fridge-magnet testHold any household magnet against the piece. Jewelry-grade 316L is austenitic, which makes it essentially non-magnetic - you may feel the faintest drag on a cold-worked clasp, and nothing more. Gold and silver are not magnetic at all. A firm pull means a ferritic steel such as 430, or a steel core hiding under a coating.Takes 5 seconds - run this first
2. The weight testPut the piece on a kitchen scale and compare the number against its physical size. This is the most powerful home test available for gold, because density differences between metals are enormous and your hand is surprisingly good at detecting them once you know what to expect.Takes 30 seconds - settles gold
3. Read the hallmark under magnificationUse a loupe or your phone camera at maximum zoom. Look for 316L, 925, 585 or 750. Genuine stamps are struck cleanly with sharp edges; counterfeits are frequently shallow, uneven in depth, or slightly soft at the outline because they were cast rather than struck.Takes 1 minute - necessary, not sufficient
4. The week-on-skin testWear it daily for seven days and watch your wrist rather than the bracelet. A green mark means copper in the alloy. Skin irritation may indicate a nickel-heavy costume alloy. Neither happens with solid 316L, titanium or high-karat gold.Takes a week - the most honest test there is
5. Inspect the wear pointsLook closely at the clasp edges, the underside of the piece and any raised detail. These are where a coating opens first, and a different colour showing through is conclusive proof of plating over base metal - the one finding no listing can argue with.Takes 1 minute - conclusive when positive

The Density Table

This is the discriminator, and it works because the numbers are far apart enough to feel in the hand rather than only measure.

Read the brass row alongside the steel row and the second blind spot appears: brass and steel are close enough in density that weight cannot separate them by hand either. That is precisely why the magnet and the weight tests belong together - the magnet catches steel, the weight catches gold, and running one without the other leaves half the possibilities open.

The magnet catches steel. The weight catches gold. Neither catches brass on its own - and brass is what most fakes are made of.

Miami Cuban - What a Link Bracelet Should Weigh

316L link bracelet with a fold-over clasp, $49.

Link bracelets are where the weight test is most revealing, because the format hides emptiness well. Hollow links look identical to solid ones in a photograph and weigh a fraction as much, which is how a very cheap chain reaches a convincing size. Pick one up and it announces itself immediately. Ours are solid 316L in both tones at $49, and the gold-tone version is PVD rather than plating - which is also why the wear-point test finds no colour change on it.

Shop Cuff and Steel

Reading Hallmarks, and How Fakes Fake Them

A stamp is a claim made by whoever struck it, and there is no enforcement behind an unbranded piece from an anonymous seller.

Two useful details. Genuine stamps on solid metal are struck under pressure, so they have crisp vertical edges; counterfeits are often cast into the mould and look slightly rounded or soft under magnification. And a piece marked GP or GF is telling you the truth about being plated or filled - those abbreviations are the most trustworthy marks in the whole table precisely because nobody fakes a worse claim.

Vintage - Nothing to Hide, Structurally

Open-back 316L steel cuffs, $49.

A solid forged cuff is the easiest object in the world to verify, and that is a design consequence rather than a marketing claim. There is no coating to wear through, no hollow section to hide lightness, and no clasp with a different alloy inside it. Run all five tests on one and every answer is the same answer - which is what you want from the cheapest piece you own as much as from the most expensive.

Shop Vintage

The Toothpaste Myth

Can Your Phone Detect Gold?

No, and the reason is worth stating precisely because the apps are convincing.

Apps that claim to identify gold use the phone's magnetometer, which is a compass sensor. It responds to magnetic fields, which means it responds to ferrous metals - iron, nickel, cheap steel. Gold is not magnetic, so the sensor is blind to it in principle rather than by lack of precision.

What a phone can genuinely tell you is whether something contains iron. That is useful information in exactly one scenario: confirming a gold-coloured piece has a steel core, which the fridge magnet does faster and for free.

Arc - The Wear-Point Test Has Nowhere to Look

One-piece forged 316L cuff, $49, bare and PVD gold-tone.

Test five is the one that catches most disappointments, and it is worth understanding what it looks for: a colour boundary at a worn edge. On bare steel there can be no such boundary, because the surface and the core are the same alloy. On the gold-tone version there is a boundary in principle - but PVD is measured in microns rather than fractions of one, so reaching it takes years rather than months, and that difference is the entire distinction between coating and plating.

Shop Arc

What Fake Feels Like

Beyond the formal tests, there is a set of impressions that experienced hands register in a second, and they are worth naming because they are learnable.

Wrong weight for the size. The most reliable one. Costume jewelry is usually too light; occasionally, when the base is zinc, it is oddly heavy in a dead way that feels different from dense metal.

Warm too quickly. Metal conducts heat away from skin. A piece that reaches body temperature almost instantly is either very thin or not metal at all under the surface.

Sound. Solid metal dropped on a hard surface rings; hollow or plated pieces click. It is a crude test and a surprisingly good one.

Clasp with no conviction. A well-made clasp closes with a definite action and stays where it is put. A vague, loose or gritty closure signals cost-cutting on the component that fails first.

Rough interior edges. Turn any piece over. Finished undersides are a reliable signal of care, because they are the surface a manufacturer skips when cutting cost.

Eros - The Clasp Question, Removed

316L cuff with cotton or leather centre, $59.

The clasp signal from the list above is worth taking seriously, because on mixed-material jewelry the clasp is frequently sourced separately and to a different standard than the body. That is also the most common place a cheap alloy hides inside an otherwise decent piece. An open cuff removes the component entirely - there is no clasp to be sourced badly, and nothing about the closure to inspect.

Shop Eros

Running the Tests on Our Own

It seems fair to state what our pieces do under each test, so the article can be checked against the products rather than only trusted.

The honest footnote is that the same five tests will not distinguish our steel from anyone else's genuine 316L, and that is exactly as it should be. The alloy is a specification, not a brand property - and if a competitor's piece passes all five, it passes.

Anchor Chain - Where Hidden Metal Usually Hides

316L box links with marine cord, $69, six colourways.

On mixed-material bracelets the tests get harder, because most of the metal is out of sight. Internal joints, connector pins and the parts of a link buried inside cord are where a cheaper alloy is easiest to hide and hardest to inspect. The only real defence is a maker who uses one specification throughout - which is why we run 316L in the links, the fold-over clasp and every internal joint rather than only where a customer can see it.

Shop Anchor Chain

The Bottom Line

Magnet first, weight second. The magnet catches cheap ferritic steel and hidden steel cores; the weight catches gold, because gold is roughly two and a half times denser than steel and unmistakable in the hand.

Neither catches brass on its own, which matters because brass is what most fake gold and fake silver is built on. Read the stamp under magnification, wear the piece for a week and watch your skin, then inspect the wear points for a colour boundary.

Skip the toothpaste, which works by abrasion and damages whatever you try it on, and ignore any app claiming your phone can detect gold - the sensor it uses is blind to gold in principle.