The best cocaine purity test is really two tests: one that identifies what is in the powder, and one that estimates how much cocaine it holds. Run them in that order and almost any kit gives you a useful answer.
“Purity testing” covers both. Identification tests are qualitative: they tell you what is there. Quantity tests are quantitative: they tell you roughly how much.
Both come in several shapes:
colour reagents in sealed glass ampoules
colour reagents in screw-cap vials
colour reagents in dropper bottles
TLC, thin layer chromatography, on a testing card
The shape matters less than you would think. What sets a good kit apart is how honest it is about its own false positives.
European cocaine is often strong. Of 1,980 lab-tested samples with a purity figure, half came in above 82%.
But 57% of 2,363 cocaine samples also held something else, most often procaine, caffeine, levamisole, phenacetin or lidocaine.
Fake cocaine is real too. Of 2,058 samples handed in as cocaine, 123 held no cocaine at all. They contained caffeine, amphetamine, ketamine, MDMA and cathinones such as 4-MMC, 2-MMC and NEP.
None of this shows in the powder. And a quantity test run on something that is not cocaine still gives you a number.
Ehrlich turns orange with procaine, the cut labs find most often. Clean cocaine leaves it pale.
Identification comes first, because every purity number depends on it.
A single-use cocaine test answers one question with the Morris reagent: is there cocaine? The multi-use cocaine and cuts kit answers the rest, with six reagents that each rule out a different family:
Robadope: amphetamine
Simon’s: methamphetamine and cathinones
Mecke: lidocaine
Ehrlich: benzocaine and procaine
Liebermann: levamisole, phenacetin, ephedrine
Morris: confirms cocaine
We recommend the multi-use kit. It catches the cuts a quantity test cannot.
Expected cocaine results from the kit booklet. Each reagent that stays blank rules out a cut. Tap to enlarge.
Common mistakes
The most common one is following outdated instructions. Older charts, still copied around the internet, say Marquis should stay clear with cocaine and that orange means amphetamine.
Real cocaine often carries secondary coca alkaloids from the leaf and from processing. Labs listed them in at least 169 of the samples above. They turn Marquis pink or peach and Liebermann orange, and they are not a cut.
Old guides also still sell Scott as a cocaine test. Scott turns just as blue with lidocaine, benzocaine and procaine, which is why we replaced it with Morris.
Use the booklet that came with your kit, and stir Morris for its full 30 seconds.
Cocaine with coca alkaloids: orange Liebermann, blue Morris, nothing else. Not a cut, just the plant.
Cocaine comes as the hydrochloride salt (the usual powder), as freebase or crack, and as coca paste, the unrefined step in between.
Salt: what every reagent and scale is made for, and gives the clearest colours.
Crack and freebase: cobalt-based cocaine reagents turn a fainter blue, so use a little more sample. Gram for gram, base holds about 12% more cocaine, so pure crack reads at the top of any scale.
Coca paste: full of other coca alkaloids and processing leftovers. Expect the identification reagents to show them and a colour quantity test to overstate. A testing card shows them as separate spots.
What about the bleach test?
The bleach test shows how fast a powder sinks through bleach. It can flag a few cuts, but it cannot tell you there is cocaine in the glass at all.
The bleach test in three frames. A trail tells you how a powder dissolves, not what it is.
Once you know it is cocaine and which cuts are out, a quantity test puts a rough number on it.
There are two ways. Colour reagents are the most affordable. TLC costs more per kit but gives the most complete result: it measures cocaine and shows every cut beside it.
How colour quantity tests work
A fixed amount of sample goes into a fixed amount of reagent. The deeper the colour, the more of whatever reacts.
That last part is the catch. The reagent cannot tell cocaine from a cut that reacts the same way, so that cut adds to the reading. Every colour quantity test has such cuts. The question is whether the maker tells you which.
Ampoules
You snap the top, drop the sample in and compare the colour with a card. There are two kinds:
One-liquid ampoules colour the whole liquid and use a card of named bands, from “very low” to “very high”, with no percentages.
Two-liquid ampoules separate into layers, and some print a percentage scale, which is a step up.
Both share the same problems:
The sample is measured with a tiny scoop or by eye, and too much pushes any result to the top.
The colours sit close together and shift with the light and the background.
Users report false positives, but the packaging does not list them.
A single ampoule says nothing about which cuts are in the powder.
A one-liquid ampoule at both ends of its scale. Fewer false positives than most, but a short step from weakest to strongest.A two-liquid ampoule a minute after the reaction. Photo from Reddit.
People who test a lot have documented all of this. One of the most careful home testers on Reddit has posted ampoule results for five years while also sending batches to labs.
They got the same near-black from a quarter of the scoop as from half.
A one-liquid ampoule developing over about a minute, with half the recommended sample. Photo from Reddit.
In 2025 one of their ampoules split into dark, clear and dark layers within seconds. Another commenter called it fine because of the colour.
The tester disagreed: one substance should colour the liquid evenly, so a split means something else is reacting. They threw that batch away.
A one-liquid ampoule that split into layers. The shade looked strong; the split said something else was reacting. Photo from Reddit.
Other users have posted a one-liquid ampoule reading “very high” on a sample whose reagents pointed to amphetamine, and one that turned purple instead of brown.
Screw-cap vials
Vial tests use the same chemistry with more control over the amounts. How they get there differs a lot.
Some vial kits from other brands ask for:
exactly 20 mg, weighed on a milligram scale you buy separately
three vials: extraction, developer and detection
a chart in 10% steps, finer than the roughly 10-point error the same makers claim
Their FAQ says lidocaine and related -caines “may still affect the result”. That is as far as it goes.
Our PRO Test for Cocaine Quantity keeps it simple:
two level microspoons of sample, about 20 mg, no scale needed
8 drops from the first vial into the second, then shake
read the top layer after 60 seconds against a 0 to 100% scale in 20% steps, or let the DrugsPRO app read the photo
The whole quantity test: two vials, a microspoon, a pipette and the scale printed on the tube.Our reference results at 25%, 50%, 75% and 100% cocaine, in daylight.The same references under lab light. The DrugsPRO app compares your photo with references taken in several kinds of light.
We also publish exactly what it reacts to, because every colour test reacts to something. Lidocaine turns our quantity test as dark as pure cocaine. Benzocaine and procaine stay pale.
Lidocaine is also one of the easiest cuts to rule out. Mecke in the identification kit turns yellow with it and stays clear with cocaine, and you run that kit before any quantity test anyway.
So the one blind spot is covered before it can matter. A false positive you know about, caught by the kit next to it, is worth more than a vague “may affect the result”.
Our quantity test on five samples, published as they came out. Lidocaine reads like pure cocaine, and Mecke in the identification kit flags it in seconds.
A TLC test separates the sample on a testing card before anything is measured. Cocaine, caffeine, levamisole or a -caine each run to their own height as a separate spot, including cuts no reagent colours.
You read the cocaine spot’s length against the printed scale, or in the DrugsPRO app, to within about 10 percentage points.
It has one known false positive, and we list it too: procaine, which runs level with cocaine. It is just as easy to rule out. Ehrlich in the identification kit turns orange with it, and a short step in the booklet filters it out.
Identification reagents plus a testing card are the best home lab you can have: what is in the powder, and roughly how much of each.
Cocaine, benzocaine, procaine and lidocaine on one card. Procaine stops level with cocaine, which is why Ehrlich comes first.
Colour is the weak point of every home test. Phone cameras, room lights and the angle of the shot all shift what you see.
Our DrugsPRO app reads the photo for you and corrects what distorts it, for both identification and quantity:
Identification reagents: tap the reaction and the app compares its colour with about 825 reference results across 15 reagents.
Quantity test: tap the white paper first, and the app corrects white balance and brightness before comparing the vial with cocaine standards from 25% to 100%, each photographed in three kinds of light.
TLC testing card: drag the four corners and the app straightens the card, correcting the camera angle, then measures each spot and turns its length into a percentage from lab-made curves, cocaine included.
It all runs on your phone. Your photo is not uploaded anywhere. We know of no other home test kit that does this.
DrugsPRO matching a reagent reaction (left) and measuring spots on a straightened testing card (right). Shown here with MDMA; cocaine works the same way.
They give a rough band once identification is done. Keep to the stated sample, read in daylight inside the stated time, and run the cocaine kit first.
Does lidocaine show up on a purity test?
On colour quantity tests, yes, and we say so for ours. It is also one of the easiest cuts to catch: Mecke turns yellow with lidocaine and stays clear with cocaine, and you run it before any quantity test anyway.
Do I need a milligram scale?
Not for the PRO quantity test, which uses two level microspoons. The TLC kit works from 40 mg, so a scale helps there.
Which test if I only buy one?
The cocaine and cuts kit. Every purity number depends on identification. Add the quantity test or TLC when you want the number.
No test result can tell you a substance is safe. No substance is 100% safe.This article is published for harm reduction purposes. It does not encourage the use of psychoactive substances and does not replace medical advice.
