Gold has one chemical habit that makes it easy to catch a fake: it ignores nitric acid. Copper, zinc, nickel, iron and silver do not. Put a drop of nitric acid on a streak of brass and it fizzes, turns green and disappears. Put the same drop on a streak of karat gold and the gold is still sitting there when you rinse it off. That difference is the acid test, and it is old enough that it gave English a phrase.
With gold trading at about $4,150 an ounce in mid-October 2026, roughly 26% below January’s record but still far above where most of the jewelry in a drawer was bought, a lot of people are asking the same question: is this real? This guide explains how the touchstone acid test works, why gold resists nitric acid at the level of electrode potentials, how to read the result, what fools it, what the karat marks and “gold filled” labels legally mean, and how to handle concentrated nitric acid without hurting yourself.
We sell nitric acid, from Nitric Acid 70% ACS Grade - Low Particle down to dilute technical grades. We do not sell test kits or pre-mixed karat acids, and nothing here turns a drop test into an assay. Where the authoritative sources stop, so do we: you will find no home recipes for 18k or 22k test acid in this article, because no standards body or gemological institute publishes one.
Last reviewed 19 October 2026 against the sources cited below. This is a practical chemistry guide, not medical, legal or regulatory advice, and not a substitute for a professional assay. How we researched it is set out at the end.
How do you tell if gold is real with acid?
You tell if gold is real with acid by rubbing a thin streak of it onto a black touchstone, putting a drop of nitric acid on the streak, and watching whether it survives. GIA’s description of the method is short: “Touchstone testing is based on the fact that 24k gold resists all but the strongest acids,” and “the purer the gold, the stronger the acid required to dissolve it.”
That gives two levels of answer. The first is a yes-or-no screen: is this gold at all, or a base metal with a yellow finish? A plain drop of nitric acid answers that, because base metal reacts loudly. GIA’s wording is that “a highly effervescent green reaction indicates base metal, and no further testing is required.”
The second answer is which karat. That needs graded test acids and reference needles of known karat, side by side on the same stone, and practice reading them. Most people who search “how to tell if gold is real” only need the first answer. Most people who sell or buy gold for a living need the second, and many of them confirm it with a lab method as well.
What is the acid test for gold?
The acid test is a comparison method: a streak of the unknown metal and streaks of known karat are put side by side on a touchstone, and acid is used to see which ones it dissolves. It is one of the oldest analytical methods still in daily use.
The touchstone itself is a fine-grained black stone. The Harvard–Cornell excavation at Sardis, the capital of ancient Lydia, found one, a “bar of very fine black stone,” in a house destroyed in the mid-sixth century BC. Theophrastus named Mount Tmolus as the source of such stones, and Pliny, quoting him, called the stone “Lydian.” Early editions of the Encyclopaedia Britannica still call it lapis Lydius. Museum materials references describe it as flint, Lydian stone or basanite: hard, black and fine enough to take a clean streak.
Acid was added to the touchstone much later, and the phrase followed it into English. The Oxford English Dictionary records “would stand the acid test” about silver in 1844 and a Dublin goldsmith writing of gold that would “resist the nitric acid test” in 1850. Merriam-Webster dates the figurative sense, a hard proof of worth, to 1854. The common story that it was coined by California Gold Rush assayers does not hold up: the silver citation is older than the Gold Rush.
How accurate is a touchstone? The Sardis Expedition notes that “with practice a jeweler can achieve better than 2–3% accuracy,” and that nineteenth-century Japanese jewelers reportedly reached about 1%. GIA itself calls the method an estimate and points to fire assay and X-ray fluorescence for precise results.
Why doesn’t nitric acid dissolve gold?
Nitric acid doesn’t dissolve gold because it is not a strong enough oxidizer: the nitrate ion can pull electrons from metals below about +0.96 volts, and gold sits at about +1.5 volts. Every metal used to fake or dilute gold sits below that line, which is why the acid sorts them so cleanly.
| Half-reaction | E° (volts) | Attacked by nitric acid? |
|---|---|---|
| Au³⁺ + 3e⁻ → Au | about +1.5 | No |
| [AuCl₄]⁻ + 3e⁻ → Au + 4Cl⁻ | +1.002 | Only with chloride present (aqua regia) |
| NO₃⁻ + 4H⁺ + 3e⁻ → NO + 2H₂O | +0.96 | The oxidizer |
| Ag⁺ + e⁻ → Ag | +0.80 | Yes |
| Cu²⁺ + 2e⁻ → Cu | +0.34 | Yes |
| Fe²⁺ + 2e⁻ → Fe | −0.44 | Yes |
| Zn²⁺ + 2e⁻ → Zn | −0.76 | Yes |
Standard reduction potentials from the LibreTexts reference table (Harvey, after Bard, Parsons and Jordan). Published values for Au³⁺/Au range from +1.498 to +1.52 V, so we give it as about +1.5 V.
Read the table from the nitrate line down. Silver, copper, iron and zinc all have potentials below +0.96 V, so nitric acid oxidizes them and carries them away as nitrates. With copper, the reaction gives off brown nitrogen dioxide and leaves copper(II) in solution. In concentrated nitrate it looks green; diluted, it turns the familiar blue. That is the green-blue you see when a brass streak fails.
Gold sits above the line, so nitric acid alone leaves it alone. The trick that dissolves gold is to add chloride. In aqua regia, which the London Bullion Market Association describes as “concentrated hydrochloric and nitric acids (in a ratio of 3:1 respectively),” chloride grabs each gold ion as it forms and holds it as tetrachloroaurate, [AuCl₄]⁻. That drops the effective potential to about +1.0 V, close enough to nitrate’s +0.96 V for the reaction to run. In the LBMA’s words, aqua regia “can dissolve gold (which neither acid can do on its own).” That is why GIA switches from nitric acid to aqua regia for the higher karats.
The acid test works because gold is the only common jewelry metal that sits above nitric acid’s oxidizing reach; anything below that line, including the copper and silver inside low-karat alloys, gets eaten first.
How do you do a gold acid test, step by step?
You do a gold acid test by rubbing a hidden spot of the piece on a touchstone, laying reference streaks beside it, applying acid for 20 to 40 seconds, rinsing, and comparing which streaks are left. This follows GIA’s bench method. Read the safety section before you open a bottle of concentrated acid.
- Set up for acid. Gloves, splash goggles and good ventilation, as GIA specifies. Work over a tray, with water and sodium bicarbonate within reach.
- Get under any plating. If you suspect the piece is not solid gold, GIA’s advice is to “make a small notch in an unobtrusive place on the jewelry to expose fresh metal.” A surface streak only samples the surface.
- Make the streak. Rub the piece on the stone to leave “a thin, but clearly visible, metal sample,” taken “from a place not clearly visible on the jewelry and away from solder joints.” Solder is a different alloy and will mislead you.
- Add references. Beside it, rub streaks from test needles of known karat, and label which is which.
- Apply the acid. Draw a line of acid across all the streaks. GIA instructors leave the 10k acid on for “approximately 20 to 40 seconds,” then neutralize and rinse.
- Read the result. If the 10k reference streak dissolves and your streak stays, your metal is finer than 10k. Move up to the next acid and repeat until your streak dissolves alongside a reference.
- Clean up. Neutralize, rinse the stone, and cap the acid. Never pour leftover acid back into the stock bottle.
What do the acid test results mean?
The acid test results mean one of three things: the streak stays (gold at or above the acid’s karat), the streak fades or dissolves (lower karat), or the streak fizzes green (base metal). The reactions below are the ones authoritative sources describe; we have marked the one that rests on trade experience rather than a published standard.
| What you streak | What nitric acid does | What it tells you |
|---|---|---|
| Karat gold, about 10k and up | The alloy metals in the streak are attacked; the gold stays visible | Gold, at or above the acid’s karat |
| Brass, copper, nickel alloys | “Highly effervescent green reaction” (GIA); brown fumes | Base metal; stop testing |
| Gold-plated or gold-filled item, surface only | Streak behaves like gold | Nothing yet: notch and re-test the core |
| Same item, notched | Core streak fizzes green | Gold layer over base metal |
| Silver | Dissolves as silver nitrate; jewelers describe a creamy, milky mark | Silver alloy (trade experience, not a published standard) |
The low-karat behavior has a metallurgical explanation. Refiners separate gold from silver by boiling the alloy in nitric acid, a process called parting. Britannica notes it only works when the gold content is low, around 25 to 30% depending on the entry. Above that, the gold shields the silver from the acid. A 10k alloy is 41.7% gold, so in a quick drop test its streak mostly survives. Alloys with much less gold are attacked, which fits GIA’s use of the 10k acid as the first threshold.
Can you test silver with nitric acid?
You can test silver with nitric acid in the sense that nitric acid dissolves silver and reacts loudly with the copper used to plate cheap silverware, but there is no published standard that turns that reaction into a fineness reading. Nitric acid dissolves silver as silver nitrate; that chemistry is textbook. What a drop of it looks like on sterling versus plate comes from jewelers’ experience. The usual description is a creamy or grey mark on solid sterling, and a green reaction when a file cut reaches copper under a silver plate.
The silver test acid you will see sold online, often called Schwerter’s solution, is potassium dichromate dissolved in nitric acid. We do not recommend it for home testing. Potassium dichromate is a hexavalent chromium compound, and IARC classifies hexavalent chromium compounds as Group 1, carcinogenic to humans. The color charts that circulate for it come from forums, not standards bodies.
For silver, trust the mark and a professional more than a drop. Under US law, “sterling” requires 925 parts per thousand silver, and a plated item may not be stamped “sterling” or “coin.” If an item is stamped sterling and fails a notch test, that is a labeling problem worth raising with the seller.
What karat is my gold?
Your gold’s karat is the number of parts of gold in 24 parts of metal by mass, so 18k is 18/24, or 75% gold. The acid test narrows the karat by elimination; the stamp tells you what the maker claims; only an assay settles it.
| Karat | Fraction | Gold by mass | Common mark | How GIA tests it |
|---|---|---|---|---|
| 24k | 24/24 | ~99.9% (fine) | 999 | Aqua regia |
| 22k | 22/24 | 91.67% | 916 | Aqua regia |
| 18k | 18/24 | 75.00% | 750 | Aqua regia |
| 14k | 14/24 | 58.33% | 585 | Measured nitric acid |
| 10k | 10/24 | 41.67% | 417 | Measured nitric acid |
| 9k | 9/24 | 37.50% | 375 | Measured nitric acid |
Percentages are n/24 arithmetic. Hallmark numbers are the legal standards in UK and Irish hallmarking law; 14k and 22k marks are rounded (585 and 916).
GIA says “measured strengths of nitric acid are used to test for 14k and lower,” and that higher karats are tested with aqua regia, “one part nitric acid and three parts hydrochloric acid,” through comparison and elimination. GIA does not publish the strength of each karat acid, and neither does any standards body we could find. The bottles sold as 10k, 14k, 18k and 22k acid are made to their suppliers’ own formulas. That is why we do not print recipes: a home-mixed acid that is slightly too strong will tell you your 14k ring is 10k, and one that is too weak will pass a fake.
What can fool the acid test?
The acid test can be fooled by anything that puts real gold on the surface and something else underneath: plating, gold-filled layers and gold-shelled cores. The acid only touches the streak, and the streak only samples what the stone can reach.
- Plating. US rules require gold electroplate to be at least 10k and at least 0.175 microns thick, and “heavy” electroplate at least 2.5 microns. A surface streak of a plated piece is a streak of real gold.
- Gold filled. A gold-filled layer must be at least 1/20 of the item’s metal weight. That is enough gold to test as karat gold until a notch reaches the core.
- Dense cores. Gold’s density is 19.3 g/cm³ and tungsten’s is 19.25, which is why tungsten is the classic core for fake bars. The acid only touches the gold shell; no authority claims the acid test can find a tungsten core.
- Gold over lower gold. GIA’s 2025 lab notes describe gold-plated gold: plating 10k gold can make X-ray fluorescence read it “as 14K or 18K gold without detecting any of the plating.” If plating can fool XRF, it can fool a surface streak.
Counterfeiters know this. In July 2026, during the price surge, a Malaysian news report quoted a gold dealer describing fakes plated thickly enough to pass both acid and magnet tests. For anything valuable, the order is: stamp, notch and acid, then a professional assay. The reference methods are fire assay (ISO 11426, a destructive cupellation method) and X-ray fluorescence (ISO 23345, non-destructive but, in that standard’s own words, not suitable for coated items).
An acid test is a screen, not a certificate. It can tell you a piece is not gold with high confidence. It cannot prove a piece is solid gold all the way through.
Do the magnet, float, vinegar and bite tests work?
The magnet, float, vinegar and bite tests can rule a few fakes out, but none of them can prove gold is real, and two of them have no authoritative support at all.
- Magnet test. Gold is diamagnetic: it is very weakly repelled by a magnet, not attracted. A piece that snaps to a magnet has iron or nickel in it. But many base metals, including brass and copper, are not magnetic either, so passing proves little.
- Float or density test. At 19.3 g/cm³ gold sinks, but so does nearly every metal. Measured carefully, density is useful, which is why professionals pair it with XRF. Tungsten matches gold’s density almost exactly.
- Vinegar test. Gold does not react with vinegar, but neither do many plated fakes in the time anyone watches. If nitric acid cannot dissolve gold, a much weaker acid cannot tell you anything nitric acid cannot. We found no authoritative source that endorses it.
- Bite test. No authoritative support. It also damages the piece and your teeth.
What do “gold filled,” “gold plated,” “vermeil” and “sterling” mean?
In the US, these terms have legal definitions in the FTC’s Jewelry Guides (16 CFR Part 23) and the National Gold and Silver Stamping Act, and knowing them tells you what an acid test should find.
| Term | What it requires | What a notch test should show |
|---|---|---|
| “14 Karat Gold,” “9 Karat Gold” | Gold alloy of that karat throughout, with the karat shown as conspicuously as the word “gold” | Gold all the way through |
| Gold Filled | Mechanically bonded layer of 10k or better, at least 1/20 of the metal weight | Gold surface, different core |
| Gold Plate / Rolled Gold Plate | Same, but at least 1/40 of the metal weight | Gold surface, different core |
| Gold Electroplate | 10k or better, at least 0.175 microns | Thin gold, base core |
| Heavy Gold Electroplate | 10k or better, at least 2.5 microns | Gold surface, base core |
| Vermeil | Sterling silver base coated with 10k or better gold, at least 2.5 microns | Gold surface, silver core |
| Sterling | At least 925/1000 silver | Silver all the way through |
Two common beliefs are out of date. First, the FTC no longer requires an alloy to be at least 10 karat to be called gold: since its 2018 revision, the Guides allow “9 Karat Gold” as long as the karat is shown just as conspicuously. The 10k floor survives only for plated, filled and vermeil layers. Second, the old “half-karat tolerance” was repealed in 1976. Under 15 U.S.C. 295, a gold article may now be no more than 3 parts per thousand below its mark, or 7 parts per thousand when the whole article is tested with its solder. For silver marked with a number, the limit is 4 parts per thousand.
Is nitric acid safe to use at home?
Nitric acid is not safe to use casually at home: it is a corrosive oxidizer that causes severe burns and gives off toxic nitrogen oxides, and its worst lung effects can appear hours after exposure. If you use it, use small amounts, real protective gear and real ventilation.
Hazards on our 70% safety data sheet: H272, may intensify fire (oxidizer), and H314, causes severe skin burns and eye damage. NIOSH lists delayed pulmonary edema among the effects of inhaling it, and a 2019 case-report review notes delayed edema “has been well described in almost all cases of nitric acid and nitrogen dioxide inhalation injury,” appearing around 8 to 12 hours later. If you breathe brown fumes and feel fine, get medical advice anyway.
- Exposure limits. OSHA’s limit is 2 ppm as an 8-hour average; NIOSH’s short-term limit is 4 ppm and its immediately-dangerous level is 25 ppm.
- Eyes and face. Splash goggles, and a face shield when you pour, as our SDS specifies.
- Gloves. For concentrated nitric acid, use neoprene or butyl. The University of Alabama in Huntsville glove chart rates nitrile as not recommended for 70% nitric acid and neoprene as good. Thin nitrile is only for incidental splashes, and should be changed at once. Our glove permeation guide explains why breakthrough, not the glove’s look, is what matters.
- Never mix it with organics. NIOSH lists alcohols, combustible materials and metal powders as incompatible. University of Washington safety guidance warns it “reacts violently” with alcohols, acetone and other organic materials. Do not wipe a spill with a solvent-soaked rag.
- Aqua regia is not stored. Mix only what you need. University safety guidance says not to seal or store it, because it releases gas and can pressurize a closed container.
- Dilute the right way. Always add acid to water, never water to acid.
- Spills. Absorb small spills with an inert material such as sand, then neutralize with sodium bicarbonate, as both our SDS and university guidance describe.
- Storage. In its original container, upright, cool, below eye level, in secondary containment, away from other acids and from solvents.
Which nitric acid should you buy?
For gold and silver work, buy the concentration your method calls for from a supplier that states the grade, and if you need repeatable results, buy ACS Grade. Here is our nitric acid range as it appears in the catalog today. Every one of them is the same chemical, CAS 7697-37-2; they differ in concentration and in how tightly impurities are specified.
| Product | Grade | Smallest pack | Typical buyer |
|---|---|---|---|
| Nitric Acid 70% ACS Grade - Low Particle | ACS Grade | 1 L | Refiners, labs, anyone who needs a specified concentrated acid |
| Nitric Acid 65% ACS Grade | ACS Grade | 1 L | Labs and jewelers who prefer slightly lower strength |
| Nitric Acid 70% Technical Grade | Technical Grade | 1 L | Industrial and refining use where trace specs matter less |
| Nitric Acid 65% Technical Grade | Technical Grade | 1 L | Industrial use |
| Nitric Acid 40% Technical Grade | Technical Grade | 1 L | Users who want a pre-diluted acid |
| Nitric Acid 25% Technical Grade · 20% · 5% | Technical Grade | 1 L | Dilute applications |
Three honest notes. First, none of these is a karat test acid: karat acids are graded formulations, and we do not sell them. Second, aqua regia for higher karats or for dissolving gold needs hydrochloric acid as well; we carry Hydrochloric Acid 37% ACS Reagent Grade, and our HCl grade guide explains the difference between grades. Third, a certificate of analysis for your lot is available on request. If you are refining rather than testing, our precious metal recovery guide covers parting and aqua regia at scale, and where to buy nitric acid covers grades, concentrations and shipping rules.
How we researched this
We read the sources below on 19 October 2026: GIA’s touchstone bench method and its 2025 lab notes on gold-plated gold, the Sardis Expedition record of the Lydian touchstone, the LBMA on aqua regia, standard reduction potential tables, the FTC Jewelry Guides and the National Gold and Silver Stamping Act, ISO 11426 and ISO 23345, the NIOSH Pocket Guide and IDLH documentation, university chemical safety sheets and glove charts, our own nitric acid safety data sheet, and Kitco and World Gold Council market data. Where sources disagree, as on gold’s exact potential and the date of the phrase, we say so. Silver reaction descriptions that rest on jewelers’ experience rather than a standard are labeled that way. Product names, grades and pack sizes come from our live catalog. The photographs in this article are illustrations, not test results.
Sources
- Bench Tip: Use the Touchstone Method for Testing Purity of Karat Gold, GIA
- The Touchstone Method for Testing Karat Gold Purity, GIA Jewelry Arts Instructors via Ganoksin
- Lab Notes: Gold-Plated Gold, Gems & Gemology, Summer 2025, GIA
- Touchstone, Archaeological Exploration of Sardis
- Aqua Regia Refining, London Bullion Market Association
- P1: Standard Reduction Potentials by Element, Chemistry LibreTexts
- Dissolving copper in nitric acid, Royal Society of Chemistry
- 16 CFR Part 23: Guides for the Jewelry, Precious Metals, and Pewter Industries, eCFR
- Guides for the Jewelry, Precious Metals, and Pewter Industries, final revision (83 FR 40667), Federal Trade Commission
- 15 U.S.C. Chapter 8: Falsely Stamped Gold or Silver, US Code
- ISO 11426:2021 Determination of gold: Cupellation method (fire assay), ISO
- NIOSH Pocket Guide: Nitric acid, CDC
- Delayed pulmonary edema after nitric acid inhalation, case report, Case Reports in Emergency Medicine (2019)
- Nitric Acid Safety Focus Sheet, University of Washington EH&S
- Glove Chemical Resistance Chart, University of Alabama in Huntsville
- Criteria for a Recommended Standard: Occupational Exposure to Hexavalent Chromium, NIOSH
- Parting, Encyclopaedia Britannica
- Gold price surge raises counterfeit gold concerns, New Straits Times (July 2026)
- Gold price chart (spot, 19 October 2026), Kitco
- The acid test, Grammarphobia (citing the Oxford English Dictionary)
- Gold Demand Trends Q1 2026: Supply, World Gold Council
- Safety Data Sheet: Nitric Acid 70%, Alliance Chemical
Frequently asked questions
Does real gold react to nitric acid?
No. Real gold does not react to nitric acid, because nitric acid is not a strong enough oxidizer to turn gold into gold ions. GIA notes that 24k gold "resists all but the strongest acids," and that higher karats are tested with aqua regia instead. In a low-karat alloy, the copper and silver in the streak can be attacked while the gold remains.
What does acid do to fake gold?
Nitric acid makes fake gold made of brass, copper or other base metals fizz and turn green as the metal dissolves into nitrates. GIA describes it as a highly effervescent green reaction and says no further testing is needed. Plated fakes only show this once a notch exposes the core.
Can gold-plated jewelry pass an acid test?
Yes. Gold-plated and gold-filled jewelry can pass an acid test if only the surface is streaked, because the surface really is gold. That is why GIA advises filing a small notch in a hidden spot first. Even X-ray fluorescence can be fooled by plating, according to GIA's 2025 lab notes.
What acid is used to test gold?
Nitric acid is used to test gold at 14k and below, and aqua regia, one part nitric acid to three parts hydrochloric acid, is used for higher karats, according to GIA. Commercial test acids labeled 10k, 14k, 18k and 22k are graded formulations whose exact strengths are not published by any standards body.
Can you test gold with vinegar?
Vinegar is not a reliable gold test. Gold does not react with vinegar, but neither do many fakes in the time it takes to watch, and we found no authoritative source that endorses the method. Nitric acid is a far stronger acid and is still only a screening test.
Does a magnet tell you if gold is real?
A magnet can only rule some fakes out. Gold is diamagnetic, so it is not attracted to a magnet, and a piece that sticks contains iron or nickel. But brass, copper and many other base metals are not magnetic either, so a piece that does not stick is not proven to be gold.
What is a touchstone made of?
A touchstone is made of a hard, fine-grained black stone such as flint, Lydian stone (lydite) or basanite. The name "Lydian stone" goes back to ancient Lydia, where archaeologists at Sardis found a touchstone in a house destroyed in the mid-sixth century BC.
Is 9 karat gold legally gold in the US?
Yes. Since the FTC revised its Jewelry Guides in 2018, a product made throughout of gold alloy may be described as gold below 10 karats, as in "9 Karat Gold," as long as the karat is shown just as conspicuously. The 10k minimum still applies to gold plate, gold filled and vermeil layers.
How accurate is the gold acid test?
The gold acid test is an estimate. The Sardis Expedition notes a practiced jeweler can achieve better than 2 to 3% accuracy on a touchstone. GIA points to fire assay (ISO 11426) and X-ray fluorescence for precise karat determination.
Where does the phrase "acid test" come from?
The phrase "acid test" comes from testing gold and silver with nitric acid. The Oxford English Dictionary records it about silver in 1844 and about gold in 1850, and Merriam-Webster dates the figurative meaning to 1854. It was not coined during the California Gold Rush.
Related Chemical Collections
About the author
Andre Taki
This article is for informational purposes only.
