The explanation is chemistry, and it is unusually simple for something that looks so improbable.

Copper. Trace amounts of it, held inside the crystal structure, in quantities that would be a rounding error anywhere else. Copper colours almost no other transparent gemstone on earth, and what it does inside a tourmaline is produce a luminosity that reads as emitted rather than reflected. The stone does not catch light so much as return more of it than you gave.

Manganese usually sits alongside the copper, and in one of its states it pushes the colour toward purple and pink. Gentle heating converts that manganese into a second state which does not absorb light in the same way, so the purple and pink fall away. The manganese itself doesn't disappear, it just stops pulling the colour toward pink, and the copper's blue and green come forward. This is the accepted and disclosed treatment of the category.

That is the whole of it. One element, present in the smallest imaginable quantity, doing something no other element does in any other stone.

There is a figure the trade likes to repeat, that Paraiba is ten thousand times rarer than diamond. We would rather not use it, because we cannot tell you where it came from.

The geography makes the point without it. Diamond is mined on four continents. The Brazilian material came from a small cluster of workings in the country’s north-east, first in Paraiba itself and later in neighbouring Rio Grande do Norte, and the original hill was worked out within a few years of being found. The trade says, only half in jest, that all the fine cut Brazilian material in existence would fit in one room.

Rarity in this category is not a matter of how few are found each year. It is a matter of a window that has already closed.

Brazil made the name. Mozambique made the stones.
Copper-bearing tourmaline was found in Nigeria around 2001, and a few years after that in the Alto Ligonha region of Mozambique’s Nampula province. For a while the trade treated the African material as an understudy: the same variety, a lesser passport.

It no longer does, and the salerooms are where you can watch the change happen. In June 2026, at Bonhams in New York, an unmounted 30.61 carat oval of Mozambique origin sold for US$1,452,000 against an estimate of US$150,000 to US$250,000. Nine times the low estimate.

The stone came from the collection of the American philanthropist Glorya Kaufman, in a plain hexagonal frame, with no maker’s name attached to it. The catalogue described it as a Paraiba-type tourmaline, the wording some laboratories and auction houses still use for copper-bearing material found outside Brazil. A room of bidders paid one and a half million dollars for a stone the catalogue had declined to call Paraiba outright. It is hard to think of a plainer statement of what this market is buying.

The laboratories arrived at the same place by a different route. Where Paraiba once described a location, the major gemmological laboratories now define it by chemistry and colour, a copper and manganese bearing tourmaline in the blue-to-green range, and record geographic origin as a separate finding on the report. The definition they work to settles the question outright: it describes a copper and manganese bearing tourmaline within a defined blue-to-green range, of a certain minimum saturation, and it applies that name whatever the geographic origin. Brazilian provenance still commands a premium. But a Mozambican stone of finer colour will outprice a muted Brazilian one, and it will do it every time.

Saturation outranks passport.

What the certificate says, and what it means
In December 2025, at Christie’s in New York, a Tiffany & Co. necklace built around a 13.54 carat Paraiba sold for US$4,223,000 against a high estimate of US$600,000. It is the world record for the category, and it works out at roughly US$311,000 a carat.

The stone carried a report from the American Gemological Laboratories confirming Brazilian origin, and recording that it had been heated by what the laboratory described as a relatively low temperature process.

That detail is worth sitting with, because it tells you what this category is actually judged on. Heating here is not an exception. It is the expectation. The Gemological Institute of America says so plainly in its published research: heat treatment of these tourmalines cannot always be detected, and the colour of any Paraiba should therefore be read as potentially modified by heat.

This does not mean treatment is irrelevant. It means Paraiba is assessed on a different axis from sapphire or ruby, where the presence or absence of heat can move value by a multiple. Here, minor heating is the norm, it is disclosed, and the market has priced its finest example accordingly.

Origin, by contrast, is not a matter of opinion. The laboratories establish it by measuring trace elements with a mass spectrometer and comparing the readings against reference collections of stones of known provenance. Where the result is ambiguous they record it as inconclusive rather than guess.

What matters, then, is the colour, and whether the paperwork tells you the truth about the stone in front of you. Ask for a report from a recognised laboratory confirming copper content, origin and treatment. Then look at the stone.

Why two carats is a large stone
Paraiba does not come large. Fine colour above three or four carats barely exists in any origin, which is why the scale of this category has to be read differently from every other coloured stone. A two carat Paraiba is a serious acquisition. In most other coloured stones, two carats is roughly where the conversation starts.

At the same Christie’s sale, a pair of Tiffany earrings set with Paraibas of 3.45 and 3.19 carats sold for US$1.3 million against an estimate of US$120,000 to US$180,000. Stones barely larger than the ones we are describing here, and ten times the low estimate.

Which is the practical form of the principle every dealer in this stone will tell you: colour intensity outranks carat weight, always. A smaller stone that holds its colour will beat a larger one that has let go of it, on price and on the eye.


The language of the colour
Laboratories describe colour with a grammar of their own, and the last word is the one that governs, because it names the primary hue. Greenish blue is a blue stone with green sitting behind it. Green-blue is still a blue stone, but the green has come forward. Cross the middle of the axis and the words change places: blue-green is a green stone carrying blue, and bluish green is green with only a trace of blue remaining.

Four terms, one axis, and the distance between the first and the last is the distance between two quite different stones. It sounds like pedantry until you are choosing between them, at which point it is precisely what you are paying for. The standard the laboratories work to accepts the whole of that range, from bluish green at one end through to greenish blue and the electric and neon blues at the other.

Notice, too, what the reports do not give you. A fancy coloured diamond is awarded a saturation grade, intense or vivid, and that grade travels with the stone and sets its price. Paraiba has no equivalent ladder. Saturation enters the definition only as a threshold, a minimum the stone must reach to carry the name at all, and beyond that the report gives you the hue and leaves the rest to your eye. Which is why every attempt to sell this stone in adjectives fails, and why it is photographed and filmed as often as it is described.

The Canturi stones
Five Paraiba tourmalines of Mozambique origin, running from 2.62 carats down to 1.54: 2.62, 2.13, 2.01, 1.76 and 1.54. Four ovals and a cushion.

Their colour sits at the blue end of that range. Some read greenish blue and some green-blue, which is to say that all five are blue stones, differing only in how far forward the green has come. Holding five inside a band that narrow is harder than finding any one of them. In a category this small and this variable, it is not the ordinary outcome of buying five stones. It is the reason for buying them together.

One further note, which surprises people who come to this stone from diamonds. Inclusions are present in nearly every Paraiba, and they are not read as a flaw here. Within a stone that generates its own light, the internal landscape tends to soften and spread that light rather than interrupt it. A perfectly clean Paraiba is not the ambition. A luminous one is.


A few questions worth asking

> Is a Brazilian stone always better than a Mozambican one?
No. Brazilian origin commands a premium and always has, but colour decides value before origin does. A finer Mozambican stone will outprice a weaker Brazilian one, which is why the shared laboratory definition rests on copper and manganese content, hue and saturation, and applies whatever the origin. You will still meet the term Paraiba-type for non-Brazilian material. It records where a stone was found rather than what it is.

> Should heating concern me?
Not in this category, provided it is disclosed. Low-temperature heating is the accepted treatment for Paraiba, and GIA’s published research notes that it cannot always be detected, which is why the colour of any stone in this category should be read as potentially heated. What should concern you is a stone sold without a report from a recognised laboratory.

> Why are the stones so small?
Because the material is. Fine Paraiba above three or four carats is close to non-existent, in any origin. Scale in this stone is measured against what the category produces, not against sapphire or emerald.

> What should I look at first?
The colour, in more than one light. Origin, treatment and clarity are context for a judgement your eye makes first. This is the rare instance where the most valuable thing about a gemstone is the thing you can see.