How long does a crystal take to form?

It sounds like it should have one answer, and almost every page online gives you one: millions of years. It's the line that ends up in product descriptions, because a big number sounds impressive. The trouble is that it's wrong often enough to be useless, and it quietly answers a different question from the one most people are asking.

Here's the honest version. Crystals grow at wildly different speeds depending on the mineral, the temperature, and how much dissolved material is available to them. Some quartz crystals have been shown to grow in a matter of hours. Some gypsum crystals grow at roughly a millimetre every two thousand years. That's a range of about a billion to one, and both ends are perfectly ordinary geology. This guide explains where any given crystal sits in that range, and why the question is harder than it looks.

Age and growth are two different questions

Before any of the numbers make sense, there's a distinction almost nobody draws.

A crystal has two separate histories. The first is its age: how long ago it formed. That's usually measured in millions or billions of years, and it's the number geologists can measure reasonably well, using radioactive decay in the crystal or the rock around it. The second is its growth time: how long the crystal actually spent growing. That one is far harder to measure, and for most minerals nobody has measured it directly at all.

These get treated as the same thing constantly, and they aren't. A crystal can be 130 million years old and still have grown in a week. It formed, and then it simply sat there, unchanged, for the rest of geological time. Age tells you when the growth happened. It tells you nothing about how long the growth took.

So when someone says amethyst takes millions of years to form, what they usually mean, whether they realise it or not, is that the amethyst is millions of years old. Those are not the same claim, and only one of them is supported.

The fast end: crystals that grow in hours

In 2020, researchers at Rice University published a study on quartz crystals from the Stewart pegmatite in southern California. Pegmatites are coarse-grained igneous bodies, and they produce some of the largest crystals on Earth, including much of the tourmaline, aquamarine and topaz that ends up in collections.

The team measured trace elements across single quartz crystals, then used those chemical profiles to work out how fast the crystal faces had advanced. The answer surprised them. A quartz crystal roughly 12 mm wide and 25 mm long had grown in a matter of hours; the core of it in under three hours, one internal zone in under four minutes. Extending those rates, they concluded that metre-scale crystals could form in days.

It's worth sitting with that, because it contradicts something most of us assume without ever examining it. A big crystal does not have to be a slow crystal. Given enough dissolved material moving past it, quartz can grow astonishingly fast.

Faster still, though it isn't a crystal at all: when lightning strikes sand, the silica melts and cools so quickly that it forms a natural glass called lechatelierite. The resulting branching tube, a fulgurite, forms in about a second. It has no crystal structure whatsoever, and that's precisely why it forms so fast. There's no time for the atoms to organise into a lattice. Speed and structure pull against each other.

The slow end: a millimetre every two thousand years

At the other extreme are the giant gypsum crystals of the Naica mine in Chihuahua, Mexico, the largest of which are over eleven metres long.

In 2011, a team led by Alexander Van Driessche and Juan Manuel García-Ruíz measured gypsum growth under conditions almost identical to the cave, using specialised interferometry. The slowest rate they recorded was 1.4 × 10⁻⁵ nanometres per second at 55°C. Translated into something you can picture, that's roughly one millimetre every two thousand years. A crystal beam a metre across would need something in the order of a million years.

Here's the part that turns the intuition upside down. Those crystals are the largest in the world because they grew so slowly. The water in the cave sat for an enormous span of time at a temperature right on the boundary between anhydrite and gypsum, around 54 to 58°C. That produced very low supersaturation, which meant hardly any new crystals started, and the few that did kept growing without competition for space or material. Slow, steady and undisturbed is the recipe for giants here.

Cave formations sit in similar territory. A typical stalactite grows at around 0.1 mm per year, roughly ten centimetres per thousand years, though fast-dripping caves can manage a few millimetres a year.

What actually sets the speed

Four things, mostly, and they explain nearly every example above.

Supersaturation is how overloaded the fluid is with dissolved material. High supersaturation means fast growth, but it also means many crystals start at once and compete, so you get lots of small ones. Low supersaturation means slow growth, few crystals, and the chance for those few to get large.

Temperature matters because growth rates generally climb steeply as things get warmer. At Naica, a single degree makes a substantial difference to the timescale.

Supply is simply whether new material keeps arriving. A crystal can only grow as fast as the fluid delivers to its surface, so circulating, refreshing fluid supports far faster growth than fluid sitting still.

Undisturbed space is the one people forget. Growing is only half of it; a crystal also has to survive. Most crystals that start to grow are broken, dissolved, buried or overgrown long before they get anywhere near collectable size.

That last point quietly explains a lot about the trade. Large, clean, undamaged specimens are rare not because growth is rare, but because survival is.

So how old is the amethyst on your shelf

Take the most common thing on the market: a Brazilian amethyst geode.

The basalt those geodes sit in belongs to the Paraná flood basalt province, erupted around 130 to 134 million years ago as South America and Africa pulled apart. Gas cavities were left behind in the rock as the lava cooled.

But the amethyst didn't form then. Those cavities were filled later, by silica-rich water moving through the basalt at low temperatures, well under 150°C. You can read the sequence in any geode you cut open: a green celadonite rim first, then agate, then colourless quartz, then the amethyst, then late calcite sitting in the middle of the cavity. Each layer is a separate episode.

So the accurate answer for a Brazilian amethyst is this. The host rock is Early Cretaceous. The amethyst is younger than the rock, formed by a later event. And how long the crystals themselves took to grow has never been measured. Anyone giving you a confident figure for that is guessing.

The diamond figure everyone repeats

You'll read almost everywhere that diamonds take 1 to 3.3 billion years to form. It's repeated so consistently that it sounds settled. It's a misreading.

What gets dated in a diamond isn't the crystal itself. It's the tiny mineral inclusions trapped inside it. Those inclusions tell you when material became sealed in, not how long the surrounding diamond spent growing. Diamonds genuinely are ancient, with some dated to around 3.5 billion years, and most form 150 to 200 km down in the mantle beneath the oldest, thickest parts of continents. But those numbers describe age and residence time. A diamond can form and then sit in the mantle for billions of years before a kimberlite eruption carries it up.

Diamonds also tend to grow in separate pulses rather than one continuous event, building up in layers as fresh carbon-bearing fluids pass through. How long the actual crystal growth takes is not well constrained. Saying otherwise sounds authoritative and isn't supported by the evidence.

What about lab-grown

Cultured quartz is grown in steel pressure vessels called autoclaves, under conditions that deliberately mimic natural hydrothermal systems: an alkaline solution, around 350°C, and very high pressure. Growth runs at roughly 0.4 to 1.0 mm per day, and a production cycle takes weeks to months. Lab diamonds take days to weeks, depending on size and method.

This is where a common assumption falls apart, and it's worth being clear about. Fast doesn't mean fake, and slow doesn't mean natural. A pegmatite crystal that grew in three hours is entirely natural. A cultured quartz crystal that took three months in a steel vessel is entirely synthetic. Growth speed tells you about the conditions a crystal experienced; it tells you nothing about whether it came out of the ground.

If you want to know whether a piece is natural, the questions that actually help are what species it is, where it came from, and whether it has been treated. Not how long it took.

The bottom line

There's no single number, and any page that gives you one is simplifying to the point of being wrong. Crystals grow across an extraordinary range of speeds: about a second for a fulgurite, hours for pegmatite quartz, days for salt in an evaporating pan, months for cultured quartz in a lab, a hundredth of a millimetre a year for a stalactite, and a millimetre every two thousand years for the giant gypsum beams at Naica. The mineral, the temperature and the supply decide where a piece lands.

And when someone tells you a crystal took millions of years to form, what they almost always mean is that it's millions of years old. That's a genuinely impressive fact on its own. It just isn't the same fact.

Understanding this doesn't make a specimen less remarkable. The amethyst on your shelf is a record of a Cretaceous lava flow, then water working through the rock long afterwards, then a slow chemistry of iron and radiation that gave it its colour. Knowing which part of that story you're actually holding is the difference between owning a pretty stone and owning a documented one. Every piece we sell is listed with its crystal system, origin, and its treatment status, stated plainly. Know what you own.

Sources: Pegmatite growth rates per Phelps, Lee and Morton, "Episodes of fast crystal growth in pegmatites," Nature Communications 11:4986 (2020). Naica gypsum growth rates per Van Driessche, García-Ruíz, Tsukamoto, Patiño-Lopez and Satoh, "Ultraslow growth rates of giant gypsum crystals," PNAS 108(38) (2011), and formation conditions per García-Ruiz et al., Geology 35(4) (2007). Amethyst geode formation and Paraná basalt ages per Juchem et al., Mineralium Deposita (2002) and Hartmann et al., International Geology Review (2013). Diamond ages and inclusion dating per the Gemological Institute of America. Speleothem growth rates and cultured quartz growth rates per standard geological and industry references. The conversion of the Naica growth rate into millimetres per two thousand years is ours, calculated from the published figure.