Magnet Test for Gold: Why TikTok Keeps Getting It Wrong

Short videos make the magnet test look like a magic trick: touch a magnet to your jewelry, and if it sticks, it’s “fake.” If it doesn’t, it’s “real.” That’s not how gold works—or how magnets work. The truth is more nuanced. Magnets can reveal certain fakes, but they can also flag real gold and miss plenty of impostors. Here’s how to use the magnet test properly, what it can (and can’t) tell you, and why TikTok keeps getting it wrong.

Why pure gold isn’t magnetic—and why that matters

Pure gold (24k, 99.9%) is diamagnetic, meaning it’s very weakly repelled by a magnetic field. You cannot feel that with a hand magnet. So a solid 24k coin or ingot will not “stick” to a magnet. This part the videos get right.

But most jewelry isn’t 24k. It’s alloyed to improve strength and color. Common alloys and their gold content are:

• 10k gold: 41.7% Au (stamped 417)
• 14k gold: 58.5% Au (stamped 585)
• 18k gold: 75.0% Au (stamped 750)

The remaining metals—copper, silver, zinc, nickel, palladium—change hardness, hue, and sometimes magnetism. Copper, silver, and zinc are not magnetic. Nickel can be weakly magnetic. Cobalt (rare in jewelry, but sometimes in low-karat white alloys) is magnetic. That’s the first complication: some legitimate gold alloys and solders can show a slight pull.

What TikTok gets wrong

• “If it sticks, it’s fake.” Not always. Springs and hidden parts are often steel. If the magnet grabs only the clasp, the chain may still be real gold.
• “If it doesn’t stick, it’s real.” Also wrong. Brass, copper, silver, and many stainless steels are non-magnetic. Gold-plated brass will pass the magnet test while being inexpensive underneath.
• “Magnets separate 10k from 18k.” They don’t. Both can be non-magnetic unless the alloy or solder contains magnetic metals. Magnet strength, part geometry, and friction further cloud the result.

What actually sticks on real jewelry

Even authentic gold pieces can have small magnetic components for function or durability:

• Clasps: Lobster and spring ring clasps contain a steel spring. A magnet will grab the clasp even if the body is 14k. Test the chain links away from the clasp.
• Pin backs and catches: Brooch pins are often tempered steel for strength.
• Earring posts: Some posts or backs (especially inexpensive replacements) are stainless steel and mildly magnetic.
• Solder joints: Low-karat or nickel-bearing solder can be slightly magnetic and will sit at seams or resized areas.

Practical example: a 14k, 2.5 mm rope chain might be non-magnetic along its length, but the lobster clasp snaps to a magnet. That’s normal.

Alloys and karats: when gold gets a little magnetic

Yellow gold alloys (10k–18k) are usually copper/silver/zinc heavy and non-magnetic. White gold is trickier:

• Nickel white gold (10–15% Ni): May show a faint attraction with a strong neodymium magnet. The pull is subtle—more of a nudge than a stick.
• Palladium white gold (10–15% Pd): Generally non-magnetic.
• Cobalt-bearing low-karat alloys: Uncommon but possible in industrial or vintage items; these can attract a magnet.

Takeaway: a whisper of attraction doesn’t prove a piece is fake, especially in low-karat white gold or at soldered spots.

How to perform the magnet test correctly

If you’re going to use a magnet, do it in a way that reduces false readings:

• Use the right magnet: A small neodymium disc (e.g., N52, ~10–20 mm diameter, 2–3 mm thick). Refrigerator magnets are too weak.
• Eliminate friction: Lay a thin tissue on a table and drag the magnet under it; or suspend the piece on thread. Sliding contact can fake a “stick.”
• Test multiple spots: Check the clasp separately, then mid-chain or the ring shank far from the head.
• Watch for “reach” rather than snap: A real steel core jumps to the magnet from a few millimeters away. A tiny wobble or pivot at the clasp suggests only a steel spring.
• Try the water trick: Float the piece in a shallow dish of water (in a plastic bag to keep it dry). Move the magnet beneath. Water reduces friction; subtle attraction becomes easier to see.

How to read the results

• Only the clasp reacts: Likely real gold with a steel spring. Proceed to other tests for karat confirmation.
• Whole chain or ring snaps strongly: Suspect a steel core or base metal. Many “gold-tone” and some plated steels behave this way.
• Very slight attraction at a seam: Could be solder. Check hallmarks and consider a density or acid test.
• No reaction anywhere: Could be solid gold, or non-magnetic base metals (brass, copper, silver). Magnet alone can’t confirm.

Where magnets are useful—and where they fail

Useful for catching:

• Steel-cored costume chains that are gold-colored.
• Magnetic stainless steel masquerading as gold.
• Clasps or findings that were swapped for base-metal parts.

Not useful for identifying:

• Gold-plated brass or copper (non-magnetic) vs. solid gold.
• Gold-filled (brass core) vs. solid gold. Both can be non-magnetic.
• Karat level (10k vs. 14k vs. 18k). Magnetism doesn’t correlate reliably.

Better tests to confirm gold

Use at least one of these in addition to the magnet:

• Hallmarks and construction: Look for 375, 417, 585, 750, 916, 999 stamps. Check if the stamp appears on both the clasp and body. Beware perfect stamps on very light, hollow-feeling pieces.
• Density (specific gravity): Weigh the item in air and in water to estimate density. Typical densities:
  • 10k yellow: ~11.5–12.5 g/cm³
  • 14k yellow: ~12.9–13.6 g/cm³
  • 18k yellow: ~15.2–15.9 g/cm³
  • Sterling silver: ~10.5 g/cm³
  • Brass: ~8.4–8.7 g/cm³
  • Stainless steel: ~7.7–8.0 g/cm³

  A 3 mm, 20″ solid 14k rope chain should feel heavy in hand; if the calculation lands near 8.5 g/cm³, it’s likely brass.

• Acid test: Scratch on a test stone and apply 10k/14k/18k acids. The line that survives indicates approximate karat. Note: this is surface-based and slightly abrasive; plating can fool it unless you expose the core.
• XRF analysis: Non-destructive and quick. Reads surface composition in percent. Deep plating can mask the core, so some jewelers make a tiny notch in an inconspicuous area when necessary.

Plating thickness and why magnets don’t help

Gold plating ranges from flash (~0.1 micron) to heavy plate (1–3 microns). Vermeil requires at least 2.5 microns over sterling silver. None of these thicknesses change magnet behavior. If the core is brass or silver, you’ll see no magnetic attraction. If the core is steel, you will. Magnet tests tell you about the core metal’s magnetic properties, not about the gold content or plating thickness.

Real-world examples

• 18k ring, nickel white gold, 2 mm band: No reaction along the shank; faint tug near the resized area. Likely nickel-bearing solder. Density test comes back ~15.3 g/cm³. Verdict: real 18k.
• “585” stamped rope chain, 4 mm, suspiciously light: No magnet attraction anywhere. Density ~8.8 g/cm³. Verdict: gold-plated brass with fake stamp.
• Chunky Cuban link chain that jumps to magnet: Strong snap from 1–2 cm away along the entire length. Verdict: steel core with gold color plating.

Bottom line

A magnet is a quick screen, not a truth serum. It can expose obvious steel-in-disguise and explain why your clasp clicks, but it cannot confirm gold or its karat. If a magnet raises questions, follow up with density, acid, or XRF testing. Use multiple clues—hallmarks, weight, construction—to make a confident call. That’s how professionals do it, and it’s why TikTok’s one-step answers keep missing the mark.