Gemstone certification, explained: what a laboratory report proves and what it never can

A dealer in Bangkok sends three photographs and a PDF, and the PDF is what you are really being asked to trust. Certified, in the coloured stone trade, means only that an independent house has looked at a stone and written down what it saw: the material, whether it grew underground or in a factory, what has been done to it since, and, where the extra fee was paid, which country the testers believe it came from. Nothing about beauty. Nothing about price. No promise at all that the stone in the photographs is the stone described on the sheet. That gap between document and parcel is where the money tends to go missing, and the rest of this page is about closing it.

Coloured stones are not graded the way diamonds are

Anyone arriving from the diamond counter brings the wrong expectations. Diamond paper grades: colour and clarity against fixed letters and numbers, a cut grade, a plotted map of every feather. Ruby and sapphire have no such scale and none is coming. GIA's Colored Stone Identification Report names species and variety, records whatever treatment the bench found, states weight and measurements, then describes the colour in the house's own words. Describes. Vivid blue in one building has never been calibrated against vivid blue in another.

Which is why two certified sapphires can sit a long way apart on price for reasons the paper barely touches. One heated, one not. One called Ceylon and one Madagascar. Or the duller of the pair simply cut better. Identity and treatment are what a document settles; the rest is judgement, and judgement is what a dealer is paid for.

Whose paper the market pays for

Lucerne first, because that is where the discipline began. Eduard Gübelin spent decades assembling stones of known provenance, and the origin work every house now does still rests on that idea of a reference collection. His laboratory has been issuing findings since 1923 and today works out of Lucerne, Hong Kong, Bangkok and New York. The ordinary document is a Gemmological Report. Better goods get a Gem Passport, which is a heavier object in every sense.

Basel is the other Swiss address. SSEF, a non-profit foundation set up in 1972, gets asked for by name in the Geneva salerooms, and a serious Burmese ruby will often arrive carrying both Swiss opinions rather than one. Consignors do not pay twice out of nerves. They pay twice because two findings that agree are worth more than one.

GIA is the name most buyers meet online, partly because it has been at this since 1931 and partly because its number is the easiest to check from a phone. One thing to know. Origin is a separate product there, printed only on the Colored Stone Identification and Origin Report, and offered for ten materials, ruby and sapphire at one end of the list, opal and jade at the other.

Bangkok is the working end of the trade. Adolf Peretti's GRS put pigeon's blood and royal blue into print in the first place, and those remain GRS terms judged against GRS criteria, whatever a listing implies. AIGS opened in 1978, runs a school beside the bench and is older in the city than anyone else. Lotus takes ruby, sapphire and spinel and not much besides, and colour-codes its covers so the answer shows before anyone opens anything: gold for a natural gem with no enhancement past cutting and polishing, silver for enhanced natural, black for synthetic or heavily treated. The Thai state has its own, GIT, created by Royal Decree at the end of 2003 and sitting on the committee that agreed the treatment wording most houses now share.

Beyond those, three more worth recognising. AGL in New York, whose Prestige Report is the American origin document, bought by IGI in January 2026 and still trading under its own name. CGL in Tokyo, met most often through its Unheated Corundum Report. HRD in Antwerp, which grades diamonds and keeps clear of geographic origin altogether. None of this is a league table. It is a list of doors that open when you knock.

Two names that get misused

Gem-A has been in London since 1908 and tests no stones. It teaches. The FGA and DGA letters after a gemmologist's name are Gem-A qualifications, which tell you something about the person holding the tweezers and nothing whatever about the stone in them. A Gem-A certificate, offered as a finding, is confusion at best.

The AGTA Gemological Testing Center is the sadder case. It ran in New York for more than a decade, shut in 2009, and its reports are still in circulation. A 2006 AGTA GTC document is genuine and was written by real gemmologists, and nobody there can now confirm it, correct it or look at the stone again. Age tells against paper much as it tells against everything else.

Checking a number

Type it into the house's own site. Not into a link that arrived with the offer, however convenient. What comes back is a summary of the original: identification, weight, measurements, treatment, sometimes an origin and a photograph. SSEF is the plainest about why any of this matters and treats the online check as the step that makes a report valid at all.

No lookup does not mean no laboratory. AGL runs none and supplies duplicates only for documents issued from 2018. GGTL has none. DSEF, in Idar-Oberstein, has none. With those the check is a telephone call and a civil question, which is how the whole trade did it for fifty years.

One trap deserves naming. EGL is not one organisation but several, separately owned and separately run in different countries, so an EGL number means nothing until it has been put to the exact entity printed on that page.

An appraisal is a different animal

Appraisals carry figures: replacement cost, or likely hammer price, in a currency, on a date, signed by somebody who will stand behind the number. No gem laboratory does this and they all say so in their conditions. A dealer who hands over a report and calls it proof of value has quietly swapped one document for another, and may not have noticed doing it. Take identity and treatment from the paper. Take value from comparable sales, or from an appraiser with no stake in the sale.

Nor is a report a guarantee. It records what one bench saw, on one morning.

Where it goes wrong

In-house certificates are everywhere and worth naming first. A dealer's own document, however handsomely printed, is a description by the seller. As a written commitment it has some use. As independent evidence it has none.

Mismatch is the commonest paperwork fraud and it needs no forgery: a real report travelling with a different stone. Scales, a gauge, the photograph, the inclusion diagram, ten minutes. Recutting causes the same trouble without anybody lying, since a recut moves the weight, alters the millimetres and can lift out the very fissure the bench recorded.

Then the words. GIA calls its documents reports rather than certificates and does so on purpose, so GIA certified in a listing means a report exists, or once existed, and you should find out which. A stone described as natural sapphire with nothing said about heat is not an unheated sapphire; the silence is the answer. Colour terms are worse again. Pigeon's blood on GRS paper does not follow a stone to Basel, and several houses decline the term altogether.

When the fee earns its keep

Charges rise with weight, so on a small commercial stone the fee can approach the value of the goods and a report becomes a courtesy. The sums change the moment treatment or origin is doing the pricing. A sapphire sold as unheated, a ruby sold as Burmese, an emerald offered as no oil or as Colombian: get current paper from a house the market knows. Against the premium being claimed the fee is nothing. At the top of the market a second opinion is ordinary practice, and where two houses disagree that disagreement is information and ought to be priced as such.

What the register records

GemsVerified verifies dealers rather than stones. Admission turns partly on certification practice. Which houses a dealer sends to. Whether documents actually travel with the goods they list. And whether treatment is stated on every stone as a positive entry, none included, rather than left blank for a buyer to fill in hopefully. That much goes on the register, so a buyer can tell before the first conversation whether a dealer certifies with laboratories the market recognises or with an office printer.

Our part ends there. The report belongs to the laboratory and describes one stone on one date. Whether the stone in the box is that stone is a matter of scales, a gauge and patience, and the price is between you and the market.