How To Tell If a Diamond Is Real: 10 Measured Checks
How to tell if a diamond is real starts with density: diamond reads 3.52 grams per cubic centimetre, cubic zirconia reads 5.6 to 6.0. Moissanite reads 3.22.
The 30-second check
How to tell if your Diamond Loose Stones and Rings is real
05. Specific gravity
Scale HighDiamond lands between 3.50 and 3.53 whatever its colour or clarity, because the figure follows the carbon lattice and nothing else. A result inside that band on a stone that also shows no doubling gives you two independent numbers in agreement.
Cubic zirconia reads 5.60 to 6.00, over 60 percent heavier than diamond for the same volume, and no cut or polish moves it. White sapphire reads 3.98 to 4.00. Leaded glass reads anywhere from 2.40 to 4.50 depending on how much lead went into the melt.
06. Diameter at one carat
Callipers HighA 1.00 ct round brilliant measures 6.4 to 6.5 mm across and roughly 4.0 mm deep. Half a carat measures about 5.1 mm and two carats about 8.1 mm. Diameter climbs as the cube root of weight, so doubling the carats adds only a quarter to the width.
A stone sold as one carat that measures 6.5 mm and weighs 1.65 ct is cubic zirconia priced as diamond. A stone at 6.5 mm weighing 0.92 ct is moissanite. Both figures take a minute to gather and neither depends on your eye.
07. Dispersion
Eye HighDiamond returns mostly white flashes with colour breaking at the edge of each facet. A dispersion of 0.044 is enough to see and not enough to dominate. Cut quality changes how much light comes back, not the colour spread.
Moissanite reads 0.104, over twice diamond, and the colour spread is the first thing anybody notices. Cubic zirconia reads 0.058 to 0.066 and throws orange and blue in broad sheets. White sapphire reads 0.018 and looks flat beside any of them.
08. Facet edge doubling
Loupe HighDiamond is singly refractive, so every facet junction reads as one clean line at any angle. Cubic zirconia is singly refractive too, which is why this check finds moissanite and not everything else.
Moissanite splits light into two rays that travel at different speeds, and 0.043 of separation doubles the back facet lines visibly at 10x. The doubling shows best near the culet and worst straight down through the table, which is exactly where most people look.
09. Facet count
Loupe HighA modern round brilliant totals 57 or 58 facets, and the junctions meet at points rather than sliding past each other. Cut quality shows in whether the 8 pavilion mains close cleanly at the culet.
A single cut carries 17 or 18 facets and a Swiss cut 33, both of which look coarse under a loupe and turn up in cheap melee. Quality cubic zirconia is cut to a full 57, so a correct count proves the cutting style and never the material.
10. Mohs hardness
Loupe HighA diamond worn daily for a decade keeps sharp facet junctions and a crisp girdle edge. Damage on a diamond shows up as a chip with a step in it, because diamond cleaves rather than abrades.
Cubic zirconia at 8.0 to 8.5 abrades against the quartz dust in ordinary household dirt, so a worn stone shows rounded junctions and a milky girdle within a year or two. A stone described as an heirloom with soft edges is not a hard stone.
11. Thermal conductivity
Eye MediumDiamond moves heat better than any metal, which is why the probe pegs at once. A top of scale reading on a combination tester with the moissanite channel silent is the strongest single instrument result available outside a laboratory.
Cubic zirconia is close to an insulator and reads at the bottom of the scale. Moissanite reads high enough to light a thermal-only probe as diamond, the most common false positive in the trade and the reason cheap testers stopped being adequate.
12. Table and depth percent
Callipers MediumGIA cut grading maps proportions to a grade, and stones in the upper grades cluster around 53 to 58 percent table with 59 to 62.5 percent total depth. Readings that match the report within a tenth of a millimetre tie the paper to the stone.
Proportions outside those bands describe a poorly cut stone, not a simulant. This check earns its place by catching a mismatch between a report and the object in your hand, which is a far more common swap than a badly made copy.
13. Girdle inscription
Loupe MediumAn inscription that returns a record matching the weight, measurements and clarity of the stone in front of you. Laboratories inscribe on request, so plenty of genuine diamonds carry nothing at all on the girdle.
An inscription that returns no record, or a record whose weight and measurements disagree with the stone. A paper report with no matching online entry is the weakest document in the trade and the easiest one to produce.
14. Setting metal mark
Loupe MediumA stone worth grading sits in marked precious metal. Fineness marks read 375, 585 or 750 for gold and 950 for platinum, and they sit beside a maker's mark on anything built for retail.
No mark at all under a large centre stone, or 925 on a piece sold as a diamond engagement ring. Sterling mounts are honest jewellery and they are not what a one carat diamond gets set into. The mount is the cheapest part of a piece to read and the part a seller forgets to change.
How to tell if a diamond is real starts with density: diamond reads 3.52 grams per cubic centimetre, cubic zirconia reads 5.6 to 6.0. Moissanite reads 3.22.
Diamond Simulants: What Changes and What Doesn’t
Four stones cover almost every diamond substitute in circulation: cubic zirconia, moissanite, white sapphire and leaded glass. Each is a different material, so each carries a different weight at the same size. That one fact sorts more stones than any optical trick.
The table sets the constants side by side. The last column is the weight of a 6.5 mm round brilliant cut in each material, worked out from the density ratio against diamond.
| Stone | Specific gravity | Refractive index | Dispersion | Mohs | Weight at 6.5 mm |
|---|---|---|---|---|---|
| Diamond | 3.52 | 2.417 | 0.044 | 10 | 1.00 ct |
| Moissanite | 3.21 to 3.22 | 2.65 to 2.69 | 0.104 | 9.25 | 0.92 ct |
| Cubic zirconia | 5.60 to 6.00 | 2.15 to 2.18 | 0.058 to 0.066 | 8.0 to 8.5 | 1.65 ct |
| White sapphire | 3.98 to 4.00 | 1.762 to 1.770 | 0.018 | 9 | 1.13 ct |
| Leaded glass | 2.40 to 4.50 | 1.50 to 1.70 | 0.010 to 0.031 | 5 to 6 | 0.70 to 1.28 ct |
Mineral constants for diamond, moissanite and corundum come from the Webmineral data sheets, and the geology.com diamond entry carries the same figures in plain prose. Trade density values for the manufactured stones sit in the Kassoy density table that appraisers keep at the bench. American law agrees with the mineralogy. The definition at 16 CFR 23.12 puts diamond hardness at 10, specific gravity at approximately 3.52 and refractive index at 2.42.
Read the table by column, not by row. Specific gravity pushes cubic zirconia away from everything else by a wide margin. Dispersion isolates moissanite. Refractive index isolates white sapphire, which looks glassy next to any of the others. Mohs hardness isolates glass, which abrades within weeks of daily wear.
Geometry is what does not change. A one carat round brilliant is cut to the same proportions in every one of these materials, because the cutting wheel does not care what it is grinding. Shape tells you nothing. Weight at a given shape tells you a great deal, which is the same arithmetic that separates a solid chain from a hollow one in the gold guide.
Lab-Grown Diamonds Are Real Diamonds
A laboratory-grown diamond is carbon in the same cubic lattice as a mined one, so every check on this page passes it. Density reads 3.52. Dispersion reads 0.044. A thermal probe calls it diamond, because it is diamond.
The Federal Trade Commission settled the wording in 2018. Until then the Jewelry Guides defined a diamond as a natural mineral, and the Commission struck that word out when it approved the revised guides. The current guides still require a clear and conspicuous disclosure for a laboratory-created stone, in wording placed beside the description rather than buried in small print.
The practical question therefore splits in two. Is this stone diamond or a simulant? The ten checks on this page answer that at a kitchen table. Was this diamond grown or mined? That is an instrument question, and no home test reaches it.
Separating grown from mined needs photoluminescence spectroscopy or a short-wave ultraviolet imaging unit, and both live inside grading laboratories. Growth markers do not show at 10x. A stone that measures 3.52 and shows no doubling has told you it is diamond and nothing whatsoever about its origin.
That distinction carries the money. Grown and mined stones of matching weight and grade trade at very different prices, and the gap has widened as production volume climbed. A seller who calls a stone diamond and stays quiet about the growth method has left out the fact that sets the price.
Why the Scratch Test Ruins Stones
A scratch test damages the stone and answers almost nothing. Hardness runs 10 for diamond, 9.25 for moissanite, 9 for sapphire and 8 to 8.5 for cubic zirconia. All four scratch window glass at Mohs 5.5, so a stone that marks a window has proved only that it is not glass.
Hardness is also not toughness. Diamond resists abrasion better than any other natural material and still splits along four cleavage planes under a sharp knock. A setter works around those planes on purpose.
The same caution applies to the breath test. A clean diamond clears fog in about a second because heat leaves it fast, and cubic zirconia holds the fog for five seconds or more. Then grease from a fingertip slows the diamond down and the gap closes. Use the result to raise suspicion, never to settle a question. Household myths fail the same way on precious metal, as the silver guide sets out at length.
Reading the Report Before You Read the Stone
A grading report outranks any bench test, and it counts for nothing until you check it against the laboratory’s own record. Paper is printable. The database is not.
Work through it in this order:
- Find the girdle inscription. Hold the stone dry at 10x with light raking across the girdle and rotate it slowly. A 20x or 30x loupe makes the characters legible once you have found them.
- Type the number into the laboratory’s own verification page. GIA calls that page Report Check and IGI calls it Verify Your Report, and both sit on the laboratory’s own domain rather than a reseller’s.
- Compare the returned measurements against your callipers. Diameter, total depth and table percentage should agree within a tenth of a millimetre.
- Compare the returned weight against your scale. The guides at 16 CFR 23.18 require a stated weight to be accurate to the last decimal place, so a stone advertised at .5 carat sits between .495 and .504.
- Treat a paper certificate with no online record as no certificate at all.
Step three is the one people skip. A report is easy to buy secondhand with a stone it was never written for, and swapping the stone underneath costs a counterfeiter nothing. The measurements are what tie the paper to the object in your hand. Branded pieces give you a second thread to pull, because a case or serial reference can be checked on its own, which is how the Cartier serial check works.
Where To Get a Diamond Graded by a Laboratory
Four routes exist and they differ in what each one can prove.
A full laboratory report from GIA or IGI is the only document that settles both questions at once: diamond or simulant, grown or mined. It takes weeks, it wants the stone unset, and the fee is hard to justify against a small stone.
A local appraiser with a combination tester settles the simulant question in minutes. The word combination is what matters. A thermal-only probe calls moissanite a diamond, because silicon carbide conducts heat well enough to trip it, and the electrical conductivity channel is what pulls the two apart.
A jeweller with a hydrostatic balance can hand you a specific gravity figure on a loose stone while you wait. That figure on its own rules out cubic zirconia, glass and white sapphire.
Legitly sits in front of all three. It reads the stone and its setting from a photograph, flags the rainbow fire, rounded facet junctions and unmarked mounts that point at a simulant, and tells you whether a piece deserves the cost of a report. It is a pre-screen, not a certificate, and it cannot see inside a stone. Use it to decide what is worth sending, the way the Van Cleef Alhambra guide applies the same step to pave set pieces.
Cut and clarity vocabulary sits alongside all of this. Federal law sets only a floor: section 23.12 bars the unqualified word diamond on a stone that has not been symmetrically fashioned with at least seventeen polished facets. Grading laboratories set their proportion bands far above that floor, and an SI1 clarity grade means inclusions a trained eye finds at 10x. A flawless looking stone at a bargain price is a warning rather than a bonus, because cubic zirconia grows clean and sells for a few dollars a carat.
Verdict: Weigh It, Measure It, Then Believe It
Two numbers outside range means walk away. One number outside range means measure again, because an unzeroed scale and a calliper held at an angle both err in the same direction.
Run the checks cheapest first. Diameter against weight catches cubic zirconia in thirty seconds with nothing but callipers and a scale. Doubling at the pavilion junctions catches moissanite under a loupe. Specific gravity catches the rest, provided the stone is loose.
A set stone limits you to the optical checks, so the mount becomes evidence. Nobody sets a genuine one carat stone into an unmarked base metal shank. Read the fineness mark first. If it reads 925 on a piece sold with a large centre stone, the seller has already named the price bracket.
Measurements for the metal around the stone sit under jewellery, and the diamond hub gathers the stone specific pages as they land. The stones sold in place of diamond have names of their own, and fake diamond names separates moissanite, cubic zirconia and glass by the numbers each one carries.
FAQ
1. How can you tell if a diamond is real?
Measure the stone across the girdle with callipers and weigh it, because density separates diamond from every common simulant before any optical test does. Legitly reads the stone and its setting from a photo and flags the fire, facet wear and mount markings that point at a simulant, though it cannot grade a stone. A one carat round brilliant measures 6.4 to 6.5 mm across. A stone at 6.5 mm weighing 1.65 carats is cubic zirconia.
2. Will a real diamond scratch glass?
Yes, and so will cubic zirconia, moissanite and white sapphire. Window glass sits at Mohs 5.5 and every common simulant is harder than that. The test rules out glass and nothing else, and it leaves a mark on whatever you dragged the stone across.
3. Do fake diamonds sparkle more than real ones?
Often they do. Dispersion measures how far a stone splits white light into colours, and diamond reads 0.044 against 0.104 for moissanite and roughly 0.060 for cubic zirconia. Under a single point light source a diamond throws mostly white flashes with colour at the edges. A stone throwing constant rainbows across the table facet is usually moissanite.
4. Can a diamond tester be wrong?
A thermal-only tester calls moissanite a diamond. Silicon carbide carries heat well enough to trip the probe, which is why testers sold since the late 1990s add an electrical conductivity channel. Ask which type the jeweller is holding before you take the reading as an answer.
5. Is a lab-grown diamond a fake diamond?
No. It is diamond by composition and by crystal structure, and the FTC Jewelry Guides define diamond without reference to how the stone formed. What the seller owes you is disclosure, because grown and mined stones of the same weight and grade sell at very different prices.
6. How wide is a one carat diamond?
A round brilliant at 1.00 carat measures 6.4 to 6.5 mm across the girdle and about 4.0 mm from table to culet. Fancy shapes spread differently, so an oval or a marquise of the same weight looks larger face up. Apply the weight against diameter check to round brilliants only.
7. Does the fog test work on diamonds?
Partly. A clean diamond clears breath fog in about a second and cubic zirconia holds it for five seconds or longer, so a slow clear is worth following up. Skin oil on the table facet drags a real diamond down to the same reading, which makes the test useful for suspicion and useless for proof.
8. What does an inscription on the girdle mean?
It is usually a grading report number, laser cut into the girdle at a size that wants 20x to read comfortably. The inscription is only as good as the record behind it, so type the number into the laboratory’s report check page and compare the measurements it returns against the stone. An inscription that returns no record is a problem, while no inscription at all is normal on smaller stones.
- law.cornell.edu www.law.cornell.edu/cfr/text/16/23.12
- law.cornell.edu www.law.cornell.edu/cfr/text/16/23.18
- ecfr.gov www.ecfr.gov/current/title-16/chapter-I/subchapter-B/part-23
- ftc.gov www.ftc.gov/news-events/news/press-releases/2018/07/ftc-approves-final-r
- webmineral.com webmineral.com/data/Diamond.shtml
- webmineral.com webmineral.com/data/Moissanite.shtml
- webmineral.com webmineral.com/data/Corundum.shtml
- geology.com geology.com/minerals/diamond.shtml
- gemsociety.org www.gemsociety.org/article/cubic-zirconia-value/
- kassoy.com kassoy.com/images/SC-Density_Jewelry%20Density%20Table.pdf
