Digital TownsCrystalsShop crystals
  1. Crystals & Minerals
  2. Buying guides
  3. Home Tests That Dont Work

Crystal Tests That Don't Work (and the Checks to Use Instead)

By Digital Towns Crystals ยท Last reviewed October 8, 2026

Short answer: the popular kitchen-table tests for "real" crystals mostly fail in one of two ways: they measure nothing about the stone (feeling its energy, a pendulum, a bite), or they measure something real but cannot separate a mineral from its imitation (scratching glass, a lighter flame, a magnet, a UV glow, salt water). Several also damage the piece. A 10x loupe, a scale and a cup of water, and honest seller disclosure answer the question far better.

Why do so many crystal tests fail?

A useful test needs a property that differs reliably between the real thing and the fake, and a way of reading it that does not depend on the tester's expectations. Most viral tests fail one condition or the other. Glass, resin and lab-grown quartz were chosen as imitations precisely because they share the obvious properties: they look clear, feel smooth, and often weigh about the same.

The overview of tests that do hold up lives in the real or fake tests guide. This page takes the other side: what each popular test actually responds to, and why that response does not answer "is it real?".

Which tests measure nothing about the stone?

The energy, vibration or pendulum test

Holding a stone and "feeling" whether it is genuine, or watching a pendulum swing over it, tests the person rather than the specimen. There is no physical property involved that a glass copy lacks, and two people holding the same stone report different results. In crystal traditions, how a stone feels to its owner may matter a great deal, but as an authenticity check this approach cannot be repeated, measured or compared, so it cannot catch a fake.

The bite or tooth test

Running a pearl along the teeth to feel grit comes from pearl testing, where a natural nacre surface does feel different from a smooth coated bead. On crystals it tells you nothing, and biting quartz (hardness 7) risks a chipped tooth, while biting a soft stone like selenite leaves marks.

"It's too perfect, so it's fake" and "it has flaws, so it's real"

Plenty of natural material is very clean: GIA notes that most amethyst has no inclusions you can see without magnification. And imitations can be made deliberately imperfect. Appearance by eye alone is a hint, not a result.

Which tests measure something real but cannot tell real from fake?

Scratching glass with the crystal

The logic seems sound: quartz is 7, glass is softer, so quartz should scratch glass. The problem is that "glass" is not one hardness. Geology.com tested items from textbook field kits and found glass anywhere from 4 to 7. A hard glass imitation can scratch a soft windowpane just as quartz does, so a scratch proves only that the crystal is harder than that particular pane. Meanwhile the test risks chipping a termination or the edge of a polished piece. The hardness testing guide explains how to do a proper scratch test, on rough material only.

Scratching the crystal itself

The reverse version, dragging a knife or key across a sphere or tower, is worse. The International Gem Society's guidance on scratch testing is never to scratch test a finished stone, even in an inconspicuous spot, because hidden fractures or internal stress can cause it to fracture or shatter. Keys and knives are not calibrated either, so you damage the piece for a vague answer.

The cold-to-the-touch test

This one has a real basis. The IGS guide to glass notes that glass feels warm to the touch, much warmer than most of the gems it may resemble. The trouble is resolution. Thermal conductivity tables at HyperPhysics put ordinary glass at about 0.8 W/m K and polystyrene at about 0.033, so plastic and glass differ by a factor of about twenty-five; the gap between glass and quartz is far smaller. What the hand can reliably detect is plastic or resin. A small stone, a warm room, or a piece that has sat in a pocket wipes out the finer difference.

The lighter or flame test

Holding a lighter to a stone is supposed to show that glass or plastic melts while real crystal does not. A short flame will not melt glass, so it cannot separate glass from quartz, and it can ruin the real thing. The IGS amethyst guide warns that amethyst is sensitive to extreme heat and that heated amethyst can turn brown, red or green at 400 to 500 degrees C. Sudden heating also cracks included or fractured crystals. Gemologists do use a heated needle (the hot point test) to detect plastic and wax, but the IGS describes it as a destructive procedure to use on finished gems only as a last resort. The heat-treated crystals guide covers what heat does to quartz color.

The magnet test

Magnets appear in two opposite myths: "real hematite is magnetic" and "if it is magnetic, it is fake". Both mislead. Geology.com states that hematite is not magnetic and should not respond to a common magnet, yet many specimens contain enough magnetite to be attracted. The same source says products sold as "magnetic hematite" are usually man-made materials that do not even share hematite's chemistry. So attraction does not prove fake, and no attraction does not prove real. The magnetic hematite guide sorts this out.

The UV glow test

Some sellers and buyers treat a glow under a black light as proof of authenticity, and no glow as proof of a fake. Geology.com is direct that fluorescence is rarely a diagnostic property; most minerals do not fluoresce and the property is unpredictable. Amethyst, for example, is listed by the IGS as inert under longwave UV. UV is a genuinely useful tool, but for specific jobs such as spotting some fillers and dyes, as the UV testing guide explains.

The water, salt water or float test

Floating a stone in salt water works for one material: amber, which has a specific gravity of about 1.08 according to LibreTexts' gemology text and so floats in strong brine, while most plastic copies sink. Every mineral sold as a crystal is far denser than salt water and sinks, so the test separates nothing among them. Soaking is also risky. Selenite is gypsum and can be etched by prolonged water contact, and porous or dyed stones can lose color or absorb salt.

"It magnifies text, so it's glass"

Both glass and quartz bend light: rose quartz has a refractive index of 1.544 to 1.553 per GIA, and the IGS puts glass between 1.470 and 1.700. How strongly a piece magnifies print depends on its curvature, so a rounded quartz point magnifies and a flat glass slab does not. The test measures shape, not material.

A streak test on a polished stone

The streak test is real mineralogy: Geology.com describes it as determining the color of a mineral in powdered form by scraping it on unglazed porcelain. It is meant for rough specimens. Dragging a polished sphere across a tile marks the sphere, and many minerals are harder than the plate and leave no streak at all.

What should you do instead?

Use non-destructive checks first, in this order:

  1. Look with a 10x loupe. Round free bubbles, swirl lines and molded edges point to glass; crystals, fluid pockets and straight color zones point to a mineral. See using a loupe.
  2. Measure specific gravity. A kitchen scale and a cup of water give a number that plastic and most resin cannot fake. The specific gravity test guide walks through it.
  3. Use UV for what it is good at: fillers, glues and certain dyes, not a yes or no on authenticity.
  4. Scratch test only rough, hidden spots with calibrated references, if at all.
  5. Check the price and the paperwork. A rare stone at a common stone's price fails before any test.
  6. Send valuable pieces to a lab. Lab-grown quartz, many treatments and high-value claims need instruments a home cannot match.

The same caution applies to stones sold with unusual claims attached, such as shungite: the useful questions are about measurable properties, density, luster and conductivity, rather than feelings.

Choosing pieces that break the home-test rules

Each of these is genuine material that a viral test would misjudge, which makes them good teaching pieces and good reminders to buy from listings that show the stone clearly and state size and weight.

  • Enhydro Quartz Skeletal Elestial Point: a 382 g, 102 by 64 mm natural quartz point with mobile fluid inclusions, listed as one air bubble and one oil bubble. "Bubbles mean fake" would condemn it; a loupe shows the bubbles sit inside liquid pockets, exactly as natural fluid inclusions do.
  • Black Obsidian Sphere: natural volcanic glass, offered at 50 and 60 mm. "It's glass, so it's fake" fails here, because obsidian is a natural glass. The right question is always whether the label is honest.
  • Hematite: offered tumbled, as a 39 mm sphere and as unpolished botryoidal pieces. The rough botryoidal form is the variant to choose if you want to see natural hematite and test its weight, since a magnet will not settle the question.
  • Rose Quartz Sphere: spheres from 60.5 to 87 mm. A polished sphere is the classic victim of the key-scratch test; measure its diameter and weight instead and check the result against quartz's density.
  • Selenite: gypsum in rough wands, spheres and hearts. A reminder that soaking tests can damage soft, water-sensitive minerals.

Digital Towns Market

Pieces that break the home-test rules

See all crystals and minerals

Frequently asked questions

Is the scratch-on-glass test a good way to check quartz?

No. Glass ranges from about 4 to 7 in hardness, so a hard glass imitation can scratch a pane just as quartz does. The test also risks chipping your crystal. Use a loupe and specific gravity first, and scratch only rough, hidden spots with calibrated references.

Will a lighter flame show if a crystal is fake?

No. A lighter cannot melt glass in a few seconds, so glass and quartz behave the same, and heat can permanently change amethyst color or crack included stones. Gemologists treat even a careful hot needle test as a destructive last resort.

Does a magnet prove hematite is real or fake?

Neither. Natural hematite is not magnetic, but many specimens contain magnetite and are attracted. Most tumbled "magnetic hematite" is a man-made material. A magnet response alone does not settle anything.

If a crystal glows under a black light, is it real?

Not necessarily. Fluorescence is unpredictable and rarely diagnostic: some specimens of a mineral glow and others do not, and many imitations, glues and dyes fluoresce too. Amethyst and most quartz show little or nothing under longwave UV.

Can I tell real crystals by how they feel in my hand?

Only roughly. Plastic and resin feel noticeably warmer and lighter than stone. Glass and quartz are much closer, and lab-grown quartz feels identical to natural. Treat touch as a prompt for a proper check.

Sources