There is a lot of confident advice about testing gold at home, and most of it overstates what the tests prove. Every method below can catch some fakes. None of them, short of an assay, confirms that a piece is what it claims to be.
Understanding what each test *rules out* — rather than what it "proves" — is the useful part.
The magnet test
What to do: hold a strong magnet — a neodymium magnet, not a fridge magnet — near the piece.
What it proves: gold is not magnetic. If the piece is pulled towards the magnet, it contains a substantial amount of a magnetic metal and is not solid gold.
Where it fails: a great many non-gold metals are also non-magnetic. Brass, copper, aluminium, lead and tungsten all pass. Passing the magnet test means only "not obviously iron or nickel".
Verdict: a fast negative test. Near-worthless as a positive one.
The density test
What to do: weigh the piece precisely. Then measure the volume of water it displaces. Density = mass ÷ volume.
What it proves: pure gold has a density of 19.32 g/cm³, far higher than most metals used to imitate it. Brass is about 8.5, lead about 11.3. A piece with a density near 8 is not gold at any purity.
Where it fails: three ways. First, alloyed gold is legitimately less dense than pure gold — a 14K piece is genuinely around 13 g/cm³, so you need to know the claimed karat to interpret the reading. Second, hollow pieces, clasps and settings distort the measurement. Third and most importantly, tungsten has a density of 19.25 g/cm³ — almost identical to gold. That is precisely why it appears in counterfeit bars, and it is why density alone cannot clear a large bar.
Verdict: the best home test, and genuinely useful on solid items. Cannot detect tungsten.
The acid test
What to do: a jeweller rubs the piece on a testing stone and applies graded acids. Different karats resist different acid strengths.
What it proves: the surface metal's approximate karat.
Where it fails: it tests only the surface. A thick gold plating over base metal passes. It also slightly damages the piece, and reading it well takes experience.
Verdict: reliable in a skilled jeweller's hands, defeated by heavy plating.
XRF analysis
What to do: a handheld X-ray fluorescence gun is pointed at the piece and reads its elemental composition in seconds.
What it proves: the actual composition of the metal, non-destructively — not just "is it gold" but what percentage of what.
Where it fails: XRF reads the outer layers, typically fractions of a millimetre. A sufficiently thick plating still misleads it, which is why high-value bars are sometimes tested at a drilled point or by ultrasound instead.
Verdict: what serious buyers use. Widely available at gold buyers and pawnbrokers, often free while you wait.
Ultrasonic and specific-gravity testing for bars
For bullion bars, where tungsten substitution is the real risk, the serious methods are ultrasonic testing — sound travels through tungsten at a measurably different speed than through gold — and precision specific-gravity measurement on calibrated equipment.
Any dealer buying bars at scale does this. Any individual buying a large bar from an unknown source should want it done.
Assay
Destructive, definitive, and reserved for scrap lots and refining. A sample is taken and chemically analysed. This is the only method that is genuinely conclusive.
What actually protects you
Testing is a fallback. The real protections are earlier in the process:
- Buy hallmarked, from a system with real enforcement
- Buy from established sellers with a reputation to lose
- Keep the invoice recording weight, purity and hallmark
- For bullion, buy recognised refiner products with serial numbers and assay cards
- Treat a large discount as information. Gold has a global price and there is no legitimate reason for a piece to be far below it. A price too good to be true is the most reliable warning sign there is.
Once you know the weight and confirmed karat, our gold price calculator will tell you exactly what the metal in it is worth today.