How to Tell If Your Gold Jewelry Is Real
I get this question a lot, usually from someone who bought a piece somewhere else and wants a second opinion, or who inherited something with no paperwork. Here's what I'd actually check, in order.
What does a gold hallmark actually tell you?#
A hallmark is a small stamped or laser-inscribed mark on the metal itself — look inside a ring band, near a clasp, or on the back of an earring post. On a genuine piece it tells you two things: the purity, and (increasingly) who's accountable for that claim.
Purity is usually shown as a three-digit number rather than "karat," because karat isn't a single global standard:
Karat | Purity | Common stamp | Notes |
|---|---|---|---|
24kt | 99.9% gold | 999 | Too soft for most jewelry; mainly coins/bars |
22kt | 91.6% gold | 916 | Common purity for Indian gold jewelry |
18kt | 75.0% gold | 750 | Popular for diamond-set pieces — durable enough for daily wear |
14kt | 58.5% gold | 585 or "14K" | Our Signature line runs 14kt |
9kt | 37.5% gold | 375 | Our Daily Luxe line runs 9kt — a common entry-level purity in the UK, lower-cost than 14kt/18kt elsewhere |
In India, BIS (Bureau of Indian Standards) hallmarking has required a six-digit alphanumeric HUID (Hallmark Unique Identification) on every hallmarked piece since June 2021. That code is genuinely useful because it's checkable — you can look it up through the BIS CARE app and see the exact registration it was issued under. A purity number alone (916, 750) is a claim; an HUID is a claim you can independently verify.
Outside India, hallmarking works differently and isn't always mandatory. The UK has a long-standing assay office system (marks from Birmingham, London, Sheffield, and Edinburgh) that is legally required above a minimum weight. The US has no mandatory hallmarking system at all — a stamped "14K" is a legal claim the seller is required to back up if challenged, but there's no equivalent to BIS or the UK assay offices verifying it at the point of stamping. That's worth knowing if you're buying from a US-based seller: the stamp is a starting point, not proof, in a way it increasingly is in India.
Can you actually test gold at home?#
To a point. A few home tests are worth knowing, mostly because they're good at catching an obvious fake, not because any of them confirm real gold on their own.
Test | How it works | What it's good for | Where it falls short |
|---|---|---|---|
Magnet test | Pure gold and its common jewelry alloys aren't magnetic; strong attraction to a magnet is a red flag | Quick way to rule out an obvious steel or iron-core fake | Passing proves nothing — plenty of non-gold metals also aren't magnetic |
Float/density test | Gold is dense (19.3 g/cm³ pure); it sinks fast and sits flat in water, while gold-plated base metal often floats or sinks slowly | A rough screen for hollow or heavily plated pieces | Several other metals are dense enough to sink similarly — this narrows things down, it doesn't confirm gold specifically |
Skin/discoloration test | Lower-purity or base-metal pieces can leave a green or black mark on skin over time | A loose, informal signal if you already own the piece | Skin chemistry, lotion, and sweat all affect this independent of gold content — too unreliable to act on alone |
Unglazed ceramic streak test | Pure gold leaves a golden-yellow streak on unglazed porcelain; most fakes leave a black or dark streak | A cheap, low-stakes screen | Leaves a small mark on the piece, and only separates "clearly fake" from "needs a real test" |
Nitric acid test | A jeweler applies acid to a small filed scratch; the reaction (or lack of one) indicates purity | Genuinely informative when done correctly | Damages the piece slightly and isn't something to attempt yourself — professional-only |
The honest trade-off: home tests are free and fast, but every single one of them is a screening tool, not a verdict. I wouldn't make a real decision — keep it, return it, get it appraised for insurance — based on a magnet or a bowl of water alone. They're good for "something's clearly off here," not for "this is definitely 916 gold."
What's the one method that actually proves it?#
XRF (X-ray fluorescence) testing, done by a jeweler or an independent testing lab. It reads the actual elemental composition of the metal in seconds, non-destructively, and gives you a real purity percentage rather than an inference. If a piece has no verifiable hallmark and you genuinely need to know, this is the step — most local jewelers (ours included) can do it on the spot, and it's usually free or low-cost as a walk-in service.
Short of that, a checkable hallmark is the next best thing — specifically one you can verify against an issuing body, not just read off the metal. That's the actual difference between "it says 916" and "it's registered as 916."
What should you check differently when buying online versus in person?#
In person, you can ask to see the hallmark under a loupe and, ideally, ask for it to be tested in front of you — any jeweler confident in their stock should have no issue with that. We do this routinely; it's a five-minute request, not an imposition.
Online, you're relying entirely on what the seller discloses and what arrives with the piece. What we'd actually check:
Does the listing state the exact purity (916, 750, 585, etc.) rather than a vague "gold-plated" or "gold-tone" euphemism?
Does the piece ship with a certificate or hallmark reference you can verify independently, not just a product description?
Is there a clear return or verification policy if an independent jeweler tests it and it doesn't match what was sold?
This is the same logic behind our own Open Pricing model — we show the gold cost, stone cost, and making charges separately, on every product page, precisely so there's nothing to take on faith. Every piece we sell carries its purity and certification detail up front, and you're welcome to get it independently tested; we'd rather you check than wonder.
Does a hallmark ever lie?#
Rarely, but it happens — counterfeit hallmarks exist, mostly on older or informally-sold pieces rather than from BIS-registered retailers, where the registration itself is the accountability layer. This is exactly why the HUID system matters more than the purity stamp alone: a fabricated "916" stamp is trivial to fake, but a fabricated HUID that also resolves correctly in BIS's registry is a different, much harder thing to pull off. If you're ever unsure, checking the HUID rather than just eyeballing the stamp is the stronger habit.
Written by
Abhay Khurana
Co-Founder, Aarvines Jewells
Before co-founding Aarvines, Abhay spent over a decade running an international growth-marketing agency working with 100+ brands. He now personally oversees pricing, certification and sourcing at Aarvines, and fields most customer questions on cost and quality directly.