How to Tell If Gold Is Real: 8 At-Home Tests That Actually Work
The fastest way to tell if gold is real is the hallmark check: real gold is stamped with a karat mark like 10K, 14K, 18K, 24K, 375, 585, 750, or 999. No single test is conclusive — combine three or four and you can rule out almost every common fake.
Quick Answers
- How can I tell if gold is real at home? Look for a karat hallmark first. If there is none or you suspect plating, run the magnet test, the ceramic scratch test, and the float test. If two or more raise red flags, the piece is not solid gold.
- How can I check gold instantly? A strong neodymium magnet held over the piece is the 10-second check. Real gold does not stick. If yours does, it is plated, alloyed with iron or nickel, or fake.
- Does vinegar detect fake gold? No. Vinegar does not react with real gold or with most common fakes. The vinegar test is unreliable. Skip it.
- What is the most accurate at-home test? A gold testing kit with nitric acid solutions for 10k, 14k, 18k, and 22k. Costs about $20. Catches almost every fake the magnet misses.
- When should I skip home tests? For any piece you think is worth $500 or more, get a professional XRF test before testing at home. Some home tests can scratch or chemically damage real gold and lower its value.
We buy gold every week at US Gold and Coin and we run these tests on hundreds of pieces. Most fakes get caught by the first two checks. Below are the eight at-home tests that actually work, in order from fastest to most conclusive, plus the three popular tests you should ignore.
The 8 At-Home Gold Tests at a Glance
| # | Test | Time | Cost | What it catches |
|---|---|---|---|---|
| 1 | Hallmark check | 30 seconds | Free | Unmarked or plated pieces |
| 2 | Magnet test | 10 seconds | $10–15 | Plated steel, iron-alloy fakes |
| 3 | Ceramic scratch test | 1 minute | $1–5 | Base metal under plating |
| 4 | Float test | 30 seconds | Free | Hollow fakes, aluminum cores |
| 5 | Skin discoloration test | Several hours | Free | Copper and base-metal alloys |
| 6 | Acid test | 5 minutes | ~$20 | Non-gold metals, identifies karat |
| 7 | Electronic gold tester | 30 seconds | $30–150 | Karat reading, surface plating |
| 8 | Density test | 5 minutes | ~$15 | Tungsten cores, all common fakes |
Hallmark Check
Real gold jewelry and bars are stamped with a karat mark that tells you the purity. This is the fastest check because it takes 30 seconds and a magnifying glass.
| Stamp | Meaning | Real gold? |
|---|---|---|
| 10K, 10KT, 417 | 41.7% pure gold | Yes |
| 14K, 14KT, 585 | 58.3% pure gold | Yes |
| 18K, 18KT, 750 | 75% pure gold | Yes |
| 22K, 22KT, 916 | 91.6% pure gold | Yes |
| 24K, 24KT, 999, 9999 | 99.9% or higher pure gold | Yes |
| GP, GEP, EP | Gold-electroplated (thin gold over base metal) | No (plated) |
| GF, RGP, 1/20 14K GF | Gold-filled (thicker gold layer, but not solid) | No (filled) |
| HGE | Heavy gold electroplate | No (plated) |
| No stamp | Either solid gold without a stamp (rare) or fake | Test further |
Pieces stamped GP, GEP, GF, RGP, or HGE are not solid gold. They have a gold layer over a base metal core. Pieces with no stamp at all could go either way. Most antique and very old pieces predate stamping requirements, so a missing stamp does not automatically mean fake. It just means you need more tests.
Watch for fake stamps
Counterfeiters know about hallmarks. A stamped 14K piece is not automatically real. We have seen plated jewelry stamped with fake karat marks at flea markets and on resale apps. Use the hallmark as a starting point, not a final answer.
Magnet Test
Pure gold is not magnetic. A real gold coin, bar, or solid jewelry piece will not stick to a magnet. This test takes 10 seconds and catches the most common fake gold problem: a thin gold layer plated over a steel or iron core.
What you need
A strong neodymium magnet, also called a rare-earth or N52 magnet. Refrigerator magnets are too weak. Neodymium magnets cost about $10 to $15 at any hardware store or on Amazon.
How to do it
- Set the gold piece on a flat, non-metal surface like a wooden table.
- Hold the magnet a quarter-inch above the piece without touching it.
- Slowly lower the magnet to make contact, then lift it straight up.
- If the gold lifts off the table, sticks to the magnet, or even tilts toward it, the piece is not solid gold.
A few things to watch for. Older white gold sometimes contains nickel, which can show very slight magnetic pull even though the piece is real. If your white gold has a 14K, 18K, 585, or 750 stamp and only shows mild pull, it is likely real white gold with high nickel content. Strong magnetic attraction with no stamp is a clear fail.
The magnet test misses two important fakes: tungsten and brass. Tungsten has nearly the same density as gold and is not magnetic, so tungsten-core counterfeits pass. Brass is also not magnetic. For these you need the next tests.
We have a full breakdown of why gold is not magnetic and what magnetic attraction means for different karats in our companion article on whether gold is magnetic.
Ceramic Scratch Test
Real gold leaves a gold-colored streak when dragged across an unglazed ceramic surface. Fake gold and gold-plated pieces leave a black or grey streak because their base metal shows through.
What you need
An unglazed ceramic plate or tile. The unglazed back of a kitchen tile works. Many gold testing kits come with a black streak stone, which is the proper professional version. Cost: about $1 to $5.
How to do it
- Hold the ceramic plate flat on a table.
- Drag the gold piece firmly across the unglazed surface, pressing hard enough to leave a streak.
- Examine the streak in good light.
- A gold-colored streak means real gold (or at least real gold on the surface). A black, grey, or no-color streak means the piece is not gold or has only a thin plating.
Float (Glass of Water) Test
Real gold is dense. Drop a piece into a glass of water and a solid gold piece will sink straight to the bottom immediately. Plated jewelry over hollow brass or aluminum often floats, sinks slowly, or hesitates.
What you need
A glass of room-temperature water. That is it.
How to do it
- Fill a glass about two-thirds with water.
- Drop the gold piece in.
- Watch what happens.
If the piece sinks fast and lands on the bottom, that is normal for real gold. Real gold has a density of 19.32 grams per cubic centimeter, almost 20 times heavier than water. If the piece floats, hovers in the middle, or sinks slowly with visible drift, the metal is much less dense than gold.
The float test catches hollow fakes and aluminum-core costume pieces. It does not catch solid brass (which is dense enough to sink) or tungsten (which is even denser than gold). Use it as one signal among several.
Skin Discoloration Test
Real gold does not react with skin. If your skin turns green or black where the metal touched it, the piece contains copper or other base metals that react with sweat. This test takes a day but it is free and surprisingly accurate.
How to do it
- Wear the piece (ring, chain, bracelet) for several hours during a warm or active day.
- After taking it off, look at the skin underneath.
- Green skin usually means copper. Black or grey skin usually means a silver or copper alloy. Either result means the piece is not solid gold.
Real solid gold (10K, 14K, 18K, 22K, or 24K) almost never causes skin discoloration. Gold is one of the least reactive metals on earth. The exception is thin gold-plated jewelry where the plating has worn through and the base copper or brass core is touching skin. In that case, the green or black mark tells you the plating is gone, even if the piece was originally real gold-over-base-metal.
Acid Test (Gold Testing Kit)
This is the most reliable at-home test. A gold testing kit costs about $20 and includes nitric acid solutions formulated for different karats (10k, 14k, 18k, and 22k). Real gold of the matching karat does not react. Anything weaker (lower karat or fake) shows a color change.
What you need
- Gold testing kit with acid solutions and a black testing stone (about $20 on Amazon)
- Nitrile gloves and eye protection
- A well-ventilated workspace (open garage or outdoors)
How to do it
- Put on gloves and eye protection. Set up over a non-reactive surface (glass, ceramic, or stone, not stainless steel or wood).
- Find an inconspicuous spot on the gold piece. The inside of a ring band or back of a clasp is ideal.
- Drag that spot firmly across the black testing stone to leave a visible gold-colored streak.
- Drop a single drop of acid (start with the lowest karat acid, usually 10k) onto the streak.
- Watch the streak for 20 to 30 seconds.
- A streak that stays gold-colored means the piece is at least that karat. A streak that fades, turns green, or disappears means the piece is below that karat purity.
Repeat with progressively stronger acids (14k, 18k, 22k) to pinpoint the exact karat. If even the 10k acid dissolves the streak, the piece is plated, low-purity gold, or not gold at all.
Electronic Gold Tester
Electronic gold testers (sometimes called gold test kits or gold pens) use a small electrochemical probe to read the karat of a piece in seconds. A consumer-grade tester costs $30 to $150. Professional models used in jewelry stores cost $400 to $2,000.
How they work
You apply a drop of test gel to the gold piece, touch the probe to the gel, and the meter reads the karat. The reaction strength tells the device how much gold is in the alloy. Most consumer testers can distinguish 10k, 14k, 18k, 22k, and 24k. Better testers also identify plating by reading just below the surface.
Limitations
Cheap testers struggle with thick plating because the probe reads only the surface. A high-quality plated piece can read as solid 14k on a $30 tester. Professional testers cost much more for a reason. For a one-time check on a single piece, an acid kit gives you better accuracy at a lower price.
Density (Specific Gravity) Test
This is the only at-home test that catches tungsten-core counterfeits, the most dangerous fake gold problem. Tungsten has a density of 19.25 grams per cubic centimeter, almost identical to gold's 19.32. The two metals look identical on a hallmark check, magnet test, ceramic scratch, and even acid test if the gold shell is thick enough. Density catches them.
What you need
- A gram scale with 0.01 gram precision (about $15 on Amazon)
- A graduated cylinder or measuring cup with milliliter markings
- A small piece of string
How to do it
- Weigh the gold piece in air. Record the weight in grams.
- Tie the piece to a string and lower it into a graduated cylinder filled with water. Measure how much the water level rises in milliliters.
- Divide the weight (grams) by the water displacement (milliliters). The result is the density in grams per cubic centimeter.
- Compare to known gold densities below.
| Material | Density (g/cm³) |
|---|---|
| 24K (pure) gold | 19.32 |
| 22K gold | 17.7–17.8 |
| 18K gold | 15.2–15.9 |
| 14K gold | 12.9–14.6 |
| 10K gold | 11.5–13.0 |
| Tungsten | 19.25 |
| Lead | 11.34 |
| Brass | 8.4–8.7 |
| Steel | 7.85 |
If your piece reads 19.32 but you suspect a tungsten core, the only way to be 100 percent sure is XRF or ultrasound testing, both of which we offer free at our office. The density test still catches every other common fake.
Three Tests You Should Skip
Three popular tests are unreliable, can damage real gold, or both. Save your time and your gold.
The vinegar test
You may have seen videos showing someone drop vinegar on gold and claim that no reaction means real gold. The truth: vinegar (acetic acid) is too weak to react with real gold OR with most common gold fakes including brass, bronze, copper alloys, and gold-plated steel. The test gives a false positive on almost every fake. Skip it.
The bite test
The Olympic-podium image of athletes biting their gold medals is theater. Real solid gold is soft enough that a hard bite leaves teeth marks, but so is lead, soft brass, and many cheap base metals. Modern gold coins and jewelry are alloyed for hardness anyway, so a 10K or 14K piece will not show teeth marks even if it is real. The bite test risks chipping a tooth and damaging the gold for no reliable answer.
The lighter test
Some social media posts claim that real gold does not change color when held to a lighter flame, while fake gold turns dark. This is half-true and dangerous. Real gold is heat-resistant and will not discolor, but most fake gold pieces (gold-plated brass, gold-filled jewelry) also do not visibly change color in a brief flame test. Worse, the heat can damage gemstones, soften delicate solder joints, and burn skin. The information you get is unreliable. Skip it.
How to Combine the Tests for a Real Answer
No single test is conclusive. The right approach is to combine three or four of them based on what you have.
For everyday jewelry
Hallmark check + magnet test + ceramic scratch test (or skin test if you do not want to mark the piece). If all three pass, you have a high-confidence read on real gold.
For unmarked pieces
Magnet test + density test + acid test. The hallmark is missing so you need to confirm karat by other means. Acid test gives you the karat range. Density confirms it.
For coins and bars
Hallmark/mint mark check + magnet test + density test. Most counterfeit coins and bars try to defeat the first two tests but fail density. If the piece passes all three, you are very likely safe. For high-value bars, finish with a professional XRF or ultrasound test.
For inherited or unknown pieces
Skip home tests entirely and bring it in. We see customers all the time who tested an inherited piece, scratched it on a ceramic plate, and lowered its resale value by hundreds of dollars. Free in-person XRF testing is faster, more accurate, and does not damage the piece.
When to Skip the Home Tests
Home tests are great for ruling out the obvious fakes (plated steel chains from a flea market, costume jewelry from an estate sale, suspicious online resale finds). They are wrong for high-value pieces.
Bring the piece in if any of these are true:
- You think the piece is worth $500 or more. A wrong home test can damage value far in excess of the testing cost.
- The piece is antique or a known maker. Hallmarks, patina, and original finish all matter for resale. Home tests can hurt all three.
- You are buying, not just identifying. Before you pay, get a professional test. Acid kits and magnets do not catch tungsten-core fakes. XRF does.
- The piece has stones. Acid will damage gemstones, especially porous stones like opals, pearls, turquoise, and emeralds.
Related Guides
Is Gold Magnetic?
A deeper look at the magnet test and why it works.
Gold Karat Guide: 10k vs 14k vs 18k vs 24k
What each karat actually means for purity, durability, and resale.
How Much Does a Gold Bar Weigh?
Standard sizes and weights from 1 gram to 400 ounces.
Does Gold Tarnish?
When real gold discolors and what it tells you.
Gold Price Calculator
Get the live value of your gold by karat and weight.
Frequently Asked Questions
How can you test gold at home?
How to check gold instantly?
How to identify fake gold?
Can vinegar detect fake gold?
Does real gold stick to a magnet?
Will real gold scratch on a ceramic plate?
How accurate is a gold testing kit?
Should I test inherited gold at home?
Not Sure If Your Gold Is Real?
Bring it in. We test every piece with a professional XRF analyzer. Free, no appointment, no obligation to sell.