To assay gold is to measure exactly how much of a piece is pure gold. “Assay gold” is gold whose purity has been tested and certified rather than simply claimed. Gold is valuable enough that buyers, sellers, and refiners all need to know its true purity.1Springer Link / OpenAlex. Modern Competent Methods for Purity Determination of Gold: Some Practical Aspects The tests range from quick checks that leave a piece untouched to methods that melt part of it, and only some see past a convincing surface.
What Gold Purity Numbers Mean
Purity is written in two ways, karats and fineness, and both say how much of an item is gold versus other metals.2Alex Lexington. Karat vs Fineness: Understanding Gold Purity Like a Professional A karat (K) counts parts of gold out of 24, so 24 karat is considered pure gold.3Wikipedia. Fineness Fineness counts parts per 1,000, so an alloy that is 75% gold is stamped “750.”4Wikipedia. Fineness Karats are used mostly for jewelry, while fineness is the standard for bars and investment coins.5Alex Lexington. Karat vs Fineness: Understanding Gold Purity Like a Professional
- 24K: about 999 fineness, essentially pure gold.6Alex Lexington. Karat vs Fineness: Understanding Gold Purity Like a Professional
- 22K: 91.67% gold, written.9167 or 916.7Alex Lexington. Karat vs Fineness: Understanding Gold Purity Like a Professional
- 18K: 75% gold, written 750.8Wikipedia. Fineness
- 14K: 58.3% gold, written 583 or 585.9Alex Lexington. Karat vs Fineness: Understanding Gold Purity Like a Professional
- .9999 fine: “four nines,” 99.99% gold.10Alex Lexington. Karat vs Fineness: Understanding Gold Purity Like a Professional
A piece may carry a karat stamp, a fineness number, or both, depending on where it was made and what it is. Converting is simple: divide the karat number by 24 and you get the gold share, so 18 divided by 24 is 0.75, or 750 parts per thousand. For 14K, 583 is the exact fraction and 585 is the rounded figure used in many markets, so either stamp is normal.
On your own piece, the stamp is usually tiny and sits inside a band or near a clasp. A number stamp alone is a maker’s claim until a tested mark backs it.
Most jewelry isn’t 24K because pure gold is soft and scratches or bends easily. Mixing in copper, silver, or zinc makes it harder and changes its color.
Fire Assay The Most Trusted Test
Fire assay melts a weighed sample so the base metals are absorbed away and the precious metal is left to weigh. It is an international standard method for measuring gold in samples from about 10% to 99.95% gold,11iTeh Standards. ISO 11426:2021 – Gold Determination by Cupellation Fire Assay Method and the recommended method for checking the fineness of jewelry alloys.12iTeh Standards. ISO 11426:2021 – Gold Determination by Cupellation Fire Assay Method
The steps go like this. A tiny scraping or drilled sample is taken, often from a hidden spot. It is weighed, melted with lead, then heated on an absorbent cup. Silver is dissolved away with acid, and the gold left over is weighed. The weight ratio gives the fineness number.
The lead and the cup do the real work. Lead melts the sample into a liquid metal bead, and when it is heated in air the lead and other base metals oxidize. The cup is made of porous material such as bone ash, and it soaks up those oxidized metals like a sponge. Gold and silver don’t oxidize at those temperatures, so they stay behind as a small bead. That is why the method is trusted: it physically separates the gold instead of estimating it from a surface signal.
Labs commonly run more than one sample from the same item and compare, since a single sample might come from an unrepresentative spot. If the item isn’t uniform, such as a soldered joint or a gold layer over another metal, the sample location changes the result, so assayers choose it with care.
The method has limits. It destroys a small piece of the sample, so it suits bars, scrap, and production checks more than a finished heirloom. Above 99.95% purity, labs switch to measuring the impurities instead and subtracting.13iTeh Standards. ISO 11426:2021 – Gold Determination by Cupellation Fire Assay Method White gold with nickel or palladium needs modified steps.14International Organization for Standardization. ISO 11426:2014 And the lead-based process leaves lead-contaminated waste that must be handled carefully.15PubMed. Recovering lead from cupel waste generated in gold analysis by Pb-Fire assay
You would actually ask for a fire assay when selling a large amount of scrap or an unbranded bar, or when disputing a dealer’s quick scan result. Assay offices also use it before applying official marks.
Quick Tests That Leave Gold Intact
In a touchstone test, the piece is rubbed on a dark stone and the streak is tested with acids.16GIA. Bench Tip #9: Use the Touchstone Method for Testing Purity in Karat Gold The purer the gold, the stronger the acid needed to dissolve it.17GIA. Bench Tip #9: Use the Touchstone Method for Testing Purity in Karat Gold So measured nitric acid tests 14K and lower, while a nitric–hydrochloric mix tests higher karats.18GIA. Bench Tip #9: Use the Touchstone Method for Testing Purity in Karat Gold
In practice, the tester leaves a streak, then puts a drop of acid on it. The streak is often compared with streaks from reference needles of known karat, which show how the unknown one behaves. If the streak stays put, the gold is likely at least as pure as that acid tests for; if it dissolves or fades, the purity is likely lower. The result is a judgment about a range, not a precise number, and it takes practice to read well.
An XRF (X-ray fluorescence) scanner shines X-rays at the metal and reads which elements it contains, giving a result in seconds. But it only reads the outer layer, about 10 to 15 microns deep in gold.19Maxsell. Myth #1: Revealing the True Depth of XRF Gold Testing Machine Accuracy That makes it poor for items where the surface differs from the inside.20PubMed Central. THE USE OF ED-XRF TECHNOLOGY TO MEASURE GOLD ASSAY AFTER THE INTRODUCTION OF ISO23345 Solder joints, clasps, and repairs are different alloys from the body of a piece, so a reading of the clasp may not match the chain.
Density testing weighs a piece in air and then in water.21The Santa Fe Symposium on Jewelry Manufacturing Technology. Assaying of Gold Jewellery—Ancient And Modern Gold’s density is about 19.3 grams per cubic centimeter, and alloys are lighter. Conductivity testers use a magnetic probe to read how well the metal conducts electricity.22LBMA. Sigmascope Eddy Current Tester Hidden alloys, holes, or foreign inserts change the reading.23LBMA. Sigmascope Eddy Current Tester
Gold isn’t magnetic, so a strong pull signals a magnetic base metal. But many fakes use non-magnetic brass or copper, so passing proves little. When you’re given a number, ask which test produced it.
How Fakes Try To Fool The Tests
Typical plating is 0.2 to 5 microns thick.24Maxsell. Myth #1: Revealing the True Depth of XRF Gold Testing Machine Accuracy An XRF scan usually sees through that to the brass or copper beneath.25WiserMachine. Gold-Filled vs Gold-Plated vs Solid Gold: The XRF Identification Protocol for Jewelers A reading that shows a lot of copper or zinc on a piece marked as high-karat gold is a flag that something is off. An unusually thick gold layer, though, could exceed the scan depth, and a piece plated with gold over a gold alloy can read better than its inside, so a good reading alone isn’t final.
It helps to know the terms. Plated pieces have a micron-thin gold skin. Gold-filled has a much thicker bonded layer, often stamped GF. Solid gold is the same purity throughout.
Tungsten is the classic filler in fake bars because its density nearly matches gold’s.26International Journal of Creative Future and Heritage (TENIAT). PENGUKURAN KETULENAN EMAS MENGGUNAKAN KAEDAH UJIAN TANPA MUSNAH: SUATU ULASAN PERBANDINGAN The two differ by only about 0.05 grams per cubic centimeter.27International Journal of Creative Future and Heritage (TENIAT). PENGUKURAN KETULENAN EMAS MENGGUNAKAN KAEDAH UJIAN TANPA MUSNAH: SUATU ULASAN PERBANDINGAN Ordinary density checks can’t tell them apart, though a precise lab system can.28IOP Conference Series: Materials Science and Engineering. Verification of the hydrostatic weighing system with existing gold purity instruments Conductivity testers, however, can distinguish tungsten fakes from genuine fine gold coins.29Gold-analytix.de. GoldScreenPen
A surface scan can read as gold because it sees only the coating. What gives a tungsten fake away is how it behaves inside, such as conductivity, sound when struck, or an ultrasound or other check that looks through the bar. Cutting or drilling would show it for certain, but that destroys a bar a buyer may want to keep intact. That’s why one test is never enough: dealers combine tests for bars and coins.
Hallmarks And Bar Standards That Prove Testing
A hallmark is an official mark applied after independent testing. An international Convention mark requires a fineness mark, a maker’s mark, and an assay office mark together.30Sheffield Assay Office. Convention Marks Items bearing it are accepted in member countries without retesting.31Hallmarking Convention. What Is Assay Gold? Purity Testing Explained The maker’s mark is registered in the country that applied the mark.32Sheffield Assay Office. Convention Marks
A real hallmark is usually a group of small marks side by side, not just one number. The assay office symbol is what shows an independent party tested the piece, so a lone “18K” without one is only the maker’s claim. Rules differ by country, and the US has no mandatory assay office, so American jewelry usually carries only a karat stamp and maker’s mark. A hallmark with an office mark is stronger than a bare stamp because an outside body vouches for it, though fakes of those marks do exist, so a suspect piece can still be tested.
For bullion, London’s market accepts only bars meeting its Good Delivery standard.33LBMA. Good Delivery Rules and Governance Those bars must be at least 995 parts per thousand gold,34LBMA. 5. London Good Delivery – Gold and Silver and the rules cover weight, purity, and markings on the bar.35LBMA. 5. London Good Delivery – Gold and Silver
Small bars and coins often come sealed with an assay card, a certificate of weight and fineness. A bar from a recognized refiner carries more weight than an unbranded one, and a broken seal may lower resale value.
Testing Extremely Pure Gold
For gold at 999.99 parts per thousand or higher, a standard method uses a mass spectrometer to count impurities.36International Organization for Standardization. ISO 5724:2023 – Jewellery and precious metals — Determination of very high purity gold — Difference method using ICP-MS Purity is then worked out by subtracting the impurities from 100%.37International Organization for Standardization. ISO 5724:2023 – Jewellery and precious metals — Determination of very high purity gold — Difference method using ICP-MS Such methods detect impurities down to about 1 part per million,38Analytik Jena. Trace Metals Analysis in High Purity Gold Using PlasmaQuant MS Elite and a laser version removes only a microscopic speck.39Advances in Transdisciplinary Engineering. Uncertainty Evaluation for Determination of Gold in High-Purity Gold by LA-ICP-MS Subtraction wins at this level because a direct measure of gold would be too coarse to show a difference between 99.99% and 99.999%, while each trace element can be measured on its own.
For an investor, four-nines and five-nines labels rely on these lab results. Ordinary shop scanners can’t confirm that last decimal, so trust comes from the refiner’s certification.
How To Get Your Own Gold Checked
Home tests can hint at authenticity, but a professional jeweler or certified lab gives the reliable answer.40World of Test. The Essentials of Gold Testing: Ensuring Precision in Precious Metal Analysis Acid kits at home can damage soft or plated surfaces. Send valuable or unmarked pieces to a jeweler with XRF or to an assay office. For bars and coins, ask for a conductivity or combined test that can catch tungsten.
Before you sell, get a non-destructive scan from someone who isn’t buying your piece, so you have a fair reference. Ask whether the offer is based on fineness and the current gold price, since that tells you how the buyer got the number. Weigh your pieces yourself first, so you know your starting weight and can check what the buyer records. Compare more than one offer, because buyers pay different shares of the gold value.
Before you buy, look for hallmarks or a refiner’s certificate, and be wary of prices well below market. For large purchases, request a test result in writing.

