5 At-Home Tests to Spot Fake Gold Before You Sell or Buy

Before reselling inherited jewellery, accepting a "vintage" piece from a private seller, or simply verifying a recent purchase, run these five tests. None replaces a professional XRF assay — but together they catch >95% of obvious fakes.

Test 1 — The hallmark check

Pure gold pieces (24K, 21K, 18K, 14K, 22K) almost always carry a stamped hallmark indicating purity:

Look on the clasp, the inside of a ring band, or the back of a pendant. Use a 10× jeweller's loupe (~$10 on Amazon) for fine detail. Stamps that say "GP" (gold-plated), "GF" (gold-filled), "1/20 12K GF" (12K gold-filled rolled to 1/20 weight), or "RGP" (rolled gold plate) are NOT solid gold — they have a thin gold layer over a base metal core.

Test 2 — The magnet test

Pure gold is NOT magnetic. Hold a strong neodymium magnet (a small one from any DIY store) close to the piece:

• No attraction at all → consistent with gold (but doesn't prove gold; brass, copper, lead also non-magnetic)
• Strong pull → contains ferrous metal (iron, nickel) → not solid gold

This test rules out steel-core "gold" pieces popular in some grey markets. It doesn't catch gold-coloured brass or copper.

Test 3 — The density test

Gold is heavy. Specific gravity: 19.3 g/cm³ for pure gold; ~15.5 g/cm³ for 21K; ~13.5 g/cm³ for 14K. Most fake metals (brass, copper, zinc alloys) sit at 7.5–8.9 g/cm³ — roughly half the density.

Method:

1. Weigh the piece on a kitchen scale, accurate to 0.1g (mass in grams = M)
2. Tie thin string around the piece, suspend in a glass of water — the water displaced equals the volume
3. Use a measuring cup: fill to 100ml, lower the piece, read the new level. Volume = new − 100ml = V cm³
4. Density = M / V

A 21K piece should land around 14.5–15.5 g/cm³. A brass fake will land near 8 g/cm³. Half the density = fake.

Test 4 — The ceramic scratch test

Buy a small unglazed ceramic tile (a "streak plate" — $5 at lapidary supply or hardware). Drag the gold piece firmly across the tile:

• Gold streak (yellowish) → consistent with gold
• Black streak → pyrite ("fool's gold") or gold-plated brass
• No streak at all → too hard; could be steel, ceramic, or other non-gold metal

This is a destructive test on the surface of the jewellery (very minor scratch) — only use on pieces you intend to sell or are confident are fake.

Test 5 — The nitric acid test (advanced)

The most reliable home test, but requires care. Nitric acid kits (~$15-30 online) include:

• Acid bottle (10N nitric acid)
• Pipette dropper
• Reference test stones

Method:

1. Lightly scratch the piece on the test stone, leaving a small gold streak
2. Drop one drop of nitric acid on the streak
3. Streak stays gold-yellow → real gold (the karat-specific acid in the kit will dissolve lower-purity gold; 18K acid leaves 24K/22K untouched)
4. Streak turns green → base metal (brass, copper alloys)
5. Streak dissolves entirely → gold-plated only (the thin gold layer dissolves once acid penetrates the plating)

Safety: wear gloves and eye protection. Work outside or in a well-ventilated area. Neutralise spilled acid with baking soda before disposal.

When to walk to a professional

If two or more of the above tests fail OR if the piece is high-value (>$500 USD), get a professional XRF assay. Most jewellers offer this for $5-20 per piece. XRF takes 30 seconds and reads exact purity to ±0.5%.

What you can do today

1. Loupe + hallmark check — 30 seconds, costs nothing
2. Magnet test — 10 seconds with a fridge magnet
3. Weigh on a kitchen scale, cross-check against our calculator for expected market value at the stated weight and karat

If the math doesn't match — if a "50g 21K" piece weighs 35g, or the seller wants 60% of spot value for a "guaranteed 24K" coin — walk away. Real gold transactions track the spot price within a tight retail spread, never miles outside it.

Bookmark our 24K page for the daily reference, and check it before any private-market transaction.