The bleach test shows you how fast a powder sinks through household bleach. That is a physical property, not an identification, so it cannot tell you what the powder is. It was a reasonable thing to do in 1976. The people who popularised it had written it off by 1980, and the cuts that matter today were not on the market yet.
The method is fixed and simple. Fill a clear shot glass with fresh bleach at room temperature, leave it until the liquid is completely still, then drop a small sample onto the surface from a clean blade and watch. Refined cocaine hesitates for ten to fifteen seconds, then trails slowly down to the bottom. That trail is the entire signal. You are watching how quickly something dissolves and how it falls, which is a property shared by a great many white powders.
The handbook that popularised it also wrote it off
David Lee’s Cocaine Handbook is where most people met this test, and it is also where the test is abandoned. In 1976, it records, the bleach test gave a buyer a reasonable idea of what an unknown sample was worth. By 1980 it could no longer confirm that cocaine was present at all. The reason it gives is that sellers learn what buyers check.
It was built for cuts you will probably never meet
The original table lists what each 1970s adulterant did in the glass. Read it as a museum piece: several different cuts produce the same quick fall, so the glass cannot tell you which one you have, and the last row was not in the book at all.
What each cut does in bleach, and whether it still matters
Cut
Behaviour in bleach
Still common?
Mannitol, lactose
Falls quickly, but trails like cocaine
Rare
Inositol
Falls first, dragging the cocaine down with it
Rare
Quinine
Stays on top, turns pink then red
Rare
Procaine, caffeine, lidocaine
Falls quickly and settles, very little trail
The most common cuts today
Levamisole
Sinks much like cocaine
Common
Fentanyl, nitazenes
Nothing visible at all
Most lethal
The cuts that actually turn up are known. Across 790 samples sold as cocaine where cocaine was detected, the substances recorded alongside it most often were procaine, then caffeine, then levamisole, followed by phenacetin. Two of the top three are local anaesthetics, which is worth knowing if you judge a sample by how much it numbs. Those are figures for other people’s samples and not a description of yours, which is the reason to test what you have rather than read what someone else had.
Methamphetamine hydrochloride dissolves readily, so there is no slow trail to interpret in the first place. The classic meth cut, MSM, is also a white crystalline solid that goes into solution, which means the glass gives you the same picture either way. Anyone judging meth by how it behaves in bleach is reading a result that the method cannot produce.
What it cannot tell you
The common cuts all fall in much the same way, so the glass cannot separate procaine from caffeine from plain sugar, and it gives you no idea how much of the sample is cut. Levamisole sinks much as cocaine does, so it keeps the glass quiet altogether. Fentanyl and the nitazenes are active in amounts far too small to change how anything moves through a liquid. That last one is rare in cocaine and still worth ten seconds, because a strip settles it outright and no amount of watching a glass ever will.
Each thing the bleach test cannot do has something cheap that can. To find out whether it is cocaine at all, a single reagent test costs less than a coffee and takes three minutes. To see what else is in it, a multi-reagent kit compares several reactions at once. To find out how strong it is, which no colour change can tell you, a quantity test gives you a number. A reagent names the substance but not the amount, and a strip catches what no colour change will, so the three answer different questions rather than replacing each other. The bleach test answers none of them, and it destroys your sample doing it.
