How to read a lab report

The document that arrives with a serious coloured stone is short, and much of it is of no consequence at all. Two or three lines carry the price. The rest describes the stone in terms the laboratory has agreed with itself. It describes it as of one particular morning, too, which matters when the report is eleven years old and the stone has been round the trade twice since. Nothing printed on the sheet asserts that the stone lying beside it is the stone the bench examined. That is the whole difficulty with laboratory paper, and most of what follows comes out of it.

The name at the top decides the rest

Whether a report repays close reading depends first on who wrote it. Findings from Gübelin, SSEF, GIA, GRS, Lotus, AGL, GIT or CGL are priced on across the trade without much argument. Those houses are known, their methods are published, and their conclusions can be tested against one another. A document from a laboratory nobody can telephone is a sheet of measurements, and measurements can be taken by anyone with a gauge.

The date underneath the name does more work than buyers credit it with. Stones are recut, re-oiled, reset and now and then quietly swapped as they move between owners, and a report describes only what was in front of the bench on the day it was issued. Reports from the AGTA Gemological Testing Center still surface in parcels. That laboratory closed in New York in 2009, so nobody is left to confirm a document, correct it, or examine the stone again.

Verification, and the houses that offer none

Where a laboratory runs an online lookup, it expects the number to be typed into its own site rather than reached through a link supplied by the seller. What comes back is a summary, and the summary is what a buyer sets against the page in hand. SSEF is blunter than the rest and treats online verification as the step that establishes a report as valid at all. That is a stronger position than most houses take, and a sensible one for a buyer to adopt generally.

The absence of a lookup settles nothing either way. Treating it as a warning sign will cause a buyer to reject perfectly good paper. AGL has never operated one, and supplies duplicates only for documents issued from 2018 onwards. GGTL offers no verification page, nor does DSEF in Idar-Oberstein, and the trade accepts the findings of both. ICA GemLab's QR verification reaches back only as far as July 2025. It should not be confused with the International Colored Gemstone Association it is affiliated with, a trade body that has never issued a report in its life.

The line that carries the money

Treatment is where the price lives, and the wording used to record it was argued over for years before it settled. No indications of heating means the bench found no evidence that the stone had been through a furnace. The grammar of that phrase is deliberate. A finding of absence is not an undertaking that nothing ever happened, and the more careful houses take trouble to keep the distinction visible.

Indications of heating, with nothing following it, means the stone was heated and there is nothing further to record. Where flux was involved, the residues healed into the fissures are graded as minor, moderate or significant, and each step along that scale takes money with it. GRS encodes the same ladder as H, H(a), H(b) and H(c), for anyone who has learned the shorthand. A phrase such as indications of heating and diffusion, or a note that beryllium was detected, describes something else entirely. Colour has been introduced from outside the crystal, whatever name the seller has chosen to sell it under.

Emerald runs on a scale of its own, covering oil and resin in surface-reaching fissures, from insignificant through minor and moderate to significant. Some houses name the filler as well. No indications of clarity enhancement on an emerald report is the equivalent of no heat on a sapphire, and roughly as common as that comparison implies.

Where a market wants the point made loudly, a whole document gets built around it. CGL in Tokyo issues an Unheated Corundum Report for that reason, and the Japanese trade pays for it. Elsewhere the working rule is duller. A treatment line that is missing, vague, or signed by a laboratory nobody has heard of means the stone should be priced as treated.

Weights, millimetres and the fraud that needs no forgery

Species and variety sit near the top of the identification block. Corundum, variety sapphire. Beryl, variety emerald. The word natural there means the crystal came out of rock rather than out of a factory, and says nothing about what was done to it afterwards. Below that come the numbers, which is the part buyers skip and certain dealers rely on them skipping: weight in carats to two decimal places, millimetre measurements, shape and cutting style.

Those numbers are the only real defence against the commonest fraud in the business, and the fraud involves no forgery at all. A genuine report is posted with a different stone. Holograms, watermarks, embossed seals, QR codes and microprint all protect the document, and none of them protect the buyer. A set of scales and a gauge will expose the substitution in a minute or two. A tenth of a carat adrift is not rounding but another stone, or the same stone recut since the report was written. Where the document carries a photograph or an inclusion diagram, the stone goes under a lamp until whatever the diagram promises has been found inside it.

Colour, described rather than measured

Colour is set down in each laboratory's own vocabulary, and there is no scale underneath any of it. That surprises buyers arriving with a diamond's colour grade in mind. Trade names such as pigeon's blood and royal blue appear only where a house maintains written criteria for them, and those criteria belong to that house alone. GRS did more than anyone to put such terms into print, and they remain GRS judgements. A padparadscha call made in one building obliges nobody in another to agree with it. A seller quoting one laboratory's colour term against a report issued by a different laboratory is quoting nothing whatever.

Lotus meets the same difficulty more directly and colour-codes the covers of its reports. Gold for a natural gem showing no enhancement beyond cutting and polishing, silver for enhanced natural material, black for synthetics and heavily treated goods. The conclusion is legible from across a room.

Origin, which is an opinion and usually an extra

A country printed on a report is an expert opinion. It is assembled from inclusion scenes and trace-element chemistry set against a reference collection, and nothing sturdier exists in this field. Houses of the first rank do occasionally disagree about the same stone, which is why important goods travel with two reports. A buyer is better served by the disagreement than by a quiet consensus.

Notice too how often no origin appears at all. The service costs extra almost everywhere and a great many buyers never ask for it. GIA prints a country on one document only, the Colored Stone Identification and Origin Report. At AIGS the source locality is a priced addition to the Full Report, though it comes as standard on the Master Gemstone Report, worth establishing before choosing a service. CGL took the work up only in November 2025, and states openly that some localities resemble one another too closely to be separated. That may be the most useful sentence any laboratory has published on the subject.

A report without an origin line is therefore not a report holding something back. Usually nobody asked for one. One word does mislead: HRD Antwerp's diamond documents use colour origin to mean natural as against treated colour, and there is no geography in it at all.

The comments field

Comments are where the bench records whatever would not fit into a box, and an experienced buyer reads them before anything else. Filled cavities are recorded there, along with coatings, stones assembled from more than one piece, notes that a stone has been repolished since some earlier document, and any colour term the house is willing to stand behind. Sellers quote the identification and the treatment lines, and almost none of them quote the comments. Read the full document rather than the summary card that travels with smaller goods. Where the online record shows a comment the paper does not, believe the record.

What no report will tell you

Nothing on the sheet indicates whether a stone has been cut well for its material, whether the colour survives a restaurant at eight in the evening, or what anybody ought to pay for it. Two stones carrying word for word identical reports can sit a long way apart in the market, and generally do. The document exists to take identity and treatment out of the argument so that judgement has somewhere clean to stand. That is a narrower job than the word certificate suggests, and a genuinely useful one.

Buying through the register

GemsVerified sits a step behind the paper. We check whether a dealer certifies with laboratories the market recognises and discloses treatment as a condition of being listed, and we verify dealers rather than individual stones. The report remains the laboratory's word about one stone on one date. It repays learning its grammar before leaning on it.