Guide · Karats, Purity & Hallmarks
How to Check Gold Purity at Home: What Works and What Doesn’t
You can't measure gold purity accurately at home. Simple tests like the magnet, vinegar and water tests can expose obvious fakes, such as steel or brass with a gold coating, but none of them can tell 22K from 18K. To know the karat, check the hallmark and invoice, or have the piece tested on an XRF machine at a jeweller or assay centre.
That still leaves home tests with a job to do. Used sensibly, they help you decide whether a piece deserves a proper test at all.
Home tests at a glance
| Test | What it can catch | What it can't tell you | Verdict |
|---|---|---|---|
| Look for stamps | Missing or odd markings | Whether the stamp is honest | Useful first step |
| Magnet | Iron or steel cores | Karat; many fake metals aren't magnetic | Rules out some fakes |
| Vinegar | Some base metals that discolour | Karat; plating can pass | Weak |
| Water float/sink | Very light, hollow or plastic items | Almost everything else | Very weak |
| Unglazed ceramic streak | Some base metals leave a dark streak | Karat; damages the piece | Not recommended |
| Skin discolouration | High copper or base metal content | Karat; varies by skin | A hint, not a test |
| Bite test | Nothing reliable | Anything | Skip it |
| Acid testing kit | Rough karat range | Exact purity; damages surface | Best left to jewellers |
What each home test really shows
Reading the stamps
Start here because it costs nothing and harms nothing. Look on the inside of rings and bangles, near clasps and on pendant loops. Indian jewellery should carry the BIS logo, a purity grade such as 22K916 and a six-character HUID. Dubai pieces usually carry the karat or fineness (22K, 916, 21K, 875, 18K, 750). Our explainer on the BIS hallmark and HUID shows what genuine Indian marks look like.
A stamp is a claim, not proof. Still, a missing stamp on new jewellery, letters that look smudged or uneven, or a mark like "GP" or "GF" (gold plated, gold filled) tells you a lot.
The magnet test
Gold isn't magnetic. If a strong magnet grabs your chain, there's iron or steel inside. That's useful. But brass, copper, aluminium and many other cheap metals aren't magnetic either, so a piece that ignores the magnet could still be fake. Clasps sometimes contain a steel spring, so test the main body, not the clasp.
The vinegar test
Some people put a few drops of white vinegar on the piece and watch for a colour change. Some base metals darken; gold doesn't. The problem is that a gold-plated item passes as long as the plating is intact. And a 14K piece, which is real gold, may show slight discolouration on its alloy. So the vinegar test can both miss fakes and worry you about genuine pieces.
The water test
Gold is very dense and sinks. Most metals sink too. So a sinking piece proves almost nothing. If something floats, it isn't solid gold, but you probably knew that already.
The ceramic streak test
Rubbing a piece on the back of an unglazed tile leaves a streak. A golden streak is said to mean gold; a black one, base metal. In practice it scratches your jewellery, and plated items with a thick layer can leave a gold-coloured streak. We'd skip it on anything you care about.
Green or black marks on skin
If a piece leaves a green or black mark on your skin, it may contain a lot of copper or other base metals. But lotions, sweat and even genuine lower-karat gold can do this. Treat it as a reason to get a proper test, not as a verdict.
Biting
The idea comes from old films: soft gold takes a tooth mark. Lead and some other metals do too, and you can chip a tooth on a solid gold coin. Leave this one alone.
Why no home test can tell 22K from 18K
22K and 18K are both real gold. They look similar, neither is magnetic, both sink, and both pass vinegar. The difference is 16.6 percentage points of gold content, and you can't see that with kitchen tools. Colour isn't reliable either, because alloy recipes vary.
Yet that difference is real money. The table below shows today's gram rates across karats in rupees and dirhams. Look at the gap between the 22K and 18K rows.
Dubai gold rate per gram by karat (₹ INR & AED)
| Karat | 1 Gram | 8 Gram | 10 Gram | 1 Tola | 1 Ounce |
|---|---|---|---|---|---|
| 24K | ₹12,268.98AED 479.60 | ₹98,152AED 3,836.80 | ₹1,22,690AED 4,796.00 | ₹1,43,103AED 5,593.96 | ₹3,81,608AED 14,917.23 |
| 22K | ₹11,238.55AED 439.32 | ₹89,908AED 3,514.56 | ₹1,12,386AED 4,393.20 | ₹1,31,084AED 5,124.14 | ₹3,49,558AED 13,664.38 |
| 21K | ₹10,735.36AED 419.65 | ₹85,883AED 3,357.20 | ₹1,07,354AED 4,196.50 | ₹1,25,215AED 4,894.72 | ₹3,33,907AED 13,052.57 |
| 18K | ₹9,201.74AED 359.70 | ₹73,614AED 2,877.60 | ₹92,017AED 3,597.00 | ₹1,07,327AED 4,195.47 | ₹2,86,206AED 11,187.92 |
| 14K | ₹7,177.46AED 280.57 | ₹57,420AED 2,244.56 | ₹71,775AED 2,805.70 | ₹83,716AED 3,272.51 | ₹2,23,244AED 8,726.70 |
International spot: USD 4,061.90 / troy oz · AED 14,917.33 · ₹3,81,611 / troy oz
Exchange rate: 1 AED = ₹25.5817 (1 USD = ₹93.95), updated 11 Nov, 1:57 PM. Rupee prices are the Dubai rate converted to INR – not the gold rate in India, which includes import duty and GST.
Here is what that gap means for one piece. Say Neha inherits a 20 g bangle she believes is 22K. Using illustrative rates of AED 295 a gram for 22K and AED 242 for 18K, its gold value would be AED 5,900 as 22K but only AED 4,840 as 18K. That's a difference of AED 1,060, roughly ₹25,000 if you assume 23.5 rupees to the dirham, on a single bangle. No home test would have spotted it.
The methods that actually work
Hallmark plus HUID check
For Indian jewellery, enter the HUID in the BIS CARE app. It shows details registered for that piece, including purity. If the app shows something different from what you were told, or no record at all for a recent purchase, take it up with the seller.
The invoice
An itemised invoice from a reputable shop, showing karat, weight, gold rate and making charges, is the second line of proof. It's also what you'll need when you sell or declare the piece.
XRF karatmeter test
An XRF machine fires X-rays at the surface and reads the metals present, giving a purity reading in a minute or two without damaging the piece. Many jewellers, refiners and hallmarking centres offer it, often for a small fee. It reads the surface, so very thick plating can still fool it in rare cases. Our guide to karatmeter (XRF) testing explains how it works and its limits.
Fire assay
This is the lab method: a small sample is melted and the gold separated out and weighed. It's the most accurate, but it destroys a little of the metal, so it's used for bars and bulk gold rather than family jewellery.
When a home test does make sense
- Sorting a box of old jewellery into "obviously costume" and "worth testing".
- Checking a cheap bargain bought from a street seller, before spending money on a lab test.
- A quick magnet check on a chain that feels oddly heavy or light.
If something fails a home test, our list of signs your gold might not be real helps you look for other clues. Plated pieces in particular can pass several tests; see how plated, filled and solid gold differ.
Buying in Dubai? Check before you pay
The easiest purity check is the one you do at the counter. Ask for the stamp to be shown, compare the shop's rate with the live 22K rate, and insist on an itemised tax invoice.
Gold in the UAE is hallmarked and assayed, with Dubai Central Laboratory as the official lab, and established Souk and mall jewellers are used to these questions. Our guide on whether Dubai gold is pure covers what to watch for. And for choosing between karats in the first place, start with 24K vs 22K gold.
Your purity-check routine
- Find and read the stamp.
- Run the magnet test on the main body of the piece.
- Check the HUID in BIS CARE for Indian hallmarked jewellery.
- Match the piece to its invoice: karat, weight, description.
- If anything is missing or doubtful, get an XRF test at a jeweller or hallmarking centre before selling, pledging or exchanging it.
Written by Sagar Yadav and checked against official sources under our editorial policy. Spotted an error? See our corrections policy.