Every toothpaste has a number that measures how much of your teeth it removes. Almost none of them print it.
The scale has existed since the 1950s. It runs from 4 to 250. Manufacturers know exactly where their product sits — and there is no requirement anywhere in the world that they tell you.
I spent a long time in a room with a machine that brushed teeth.
Not real mouths. Extracted human teeth, set in resin blocks, mounted under a mechanical arm that held a standard toothbrush at fixed pressure and ran it back and forth fifteen hundred times. Same brush. Same pressure. Same stroke count, every time. The only thing that changed was the paste.
Afterwards we measured how much of the tooth was gone.
That is, more or less, the entire test. It has a name — relative dentin abrasivity, or RDA — and it has been the industry standard since the middle of the last century. Every major toothpaste on the shelf has been through some version of it.
What I want to tell you about is not the test. It's what happened to the results.
I'm not going to tell you that toothpaste is dangerous, or that the industry is hiding a scandal. Neither is true. What I am going to tell you is narrower and, I think, more useful: there is a specific, measurable property of the product you put in your mouth twice a day, it varies enormously between products, it is not printed on any of them, and understanding it changes what you buy. Everything below is published.
What the scale actually measures
Here is the part most people find surprising: the test isn't run on enamel.
It's run on dentin — the softer, yellower layer underneath. Dentin is used because it wears measurably faster, which makes it a more sensitive instrument. A paste that barely marks enamel will leave a readable trace on dentin, and that trace is what gets scored.
The scale is anchored to a reference material assigned a value of 100. A paste causing half that wear scores 50. A paste causing half again as much scores 150.
The scale, in full
Relative dentin abrasivity · ISO 11609
The band names are industry convention rather than regulation — the Journal of the American Dental Association notes they were proposed somewhat arbitrarily. The 250 ceiling is a formal ISO limit. A companion enamel measure, REA, is capped at 40.
For reference points at the bottom of the scale: water scores 4. Plain baking soda scores 7 — one of the softest abrasives used in oral care, with a hardness roughly equal to dentin itself and lower than enamel.
At the top, ISO sets a ceiling of 250.
That is a sixty-fold spread between the gentlest thing you could put on a brush and the maximum the standard permits. Both ends are legal. Both are sold. Neither is labelled.
The finding that should have ended the argument in 1948
In the late 1940s, two researchers named Kitchin and Robinson ran what is still one of the most quietly important experiments in the field. They wanted to know how much abrasivity a toothpaste actually needs in order to clean.
The answer: over ninety per cent of stain removal was achieved below a very low wear threshold — less than a millimetre of dentin loss per hundred thousand brush strokes, measured on an automated brushing simulator.
Read that again, because the implication is the whole story. Almost all of the cleaning happens at the gentle end. Everything above that point is buying a fractional cosmetic gain in exchange for material that does not come back.
Which raises an obvious question. If ninety per cent of the benefit is available at the bottom of the scale, why does the top of the scale exist at all?
There are only two ways to whiten a tooth
This is the part that reframed the category for me, and it comes from the American Dental Association rather than from me.
Whitening happens by one of exactly two mechanisms. Either a peroxide compound bleaches the tooth chemically, or an abrasive or polishing agent physically removes surface stain.
That's the list. There is no third option.
So when a product markets itself as a whitener and contains no peroxide, the mechanism is not in question. It is removing material. Not because the manufacturer is being deceptive — because that is the only remaining way it could work.
Charcoal is the clearest example. Its stain removal is mechanical. And there's a second effect worth knowing about: charcoal temporarily darkens the whole mouth, which raises the contrast against the teeth. Researchers have specifically noted this can mislead consumers about how much whitening has actually occurred.
Some of the visible result is contrast, not colour.
More nuanced than the popular version
Charcoal pastes are not uniformly abrasive. A 2023 study in the International Journal of Dental Hygiene tested twelve of them and found RDA values from 24 to 166, and enamel values from 0 to 14. The authors concluded these did not differ significantly from conventional toothpastes already on the market. One charcoal-only formulation scored 27, with enamel wear too low to measure.
The real problem is that nobody has established they work, or that they're safe. A literature review in the Journal of the American Dental Association found insufficient clinical and laboratory data to substantiate the safety and efficacy claims being made for charcoal dentifrices, and advised clinicians to caution patients about unproven claims.
No charcoal toothpaste has ever earned the ADA Seal of Acceptance.
And charcoal adsorbs fluoride. Activated charcoal is a general-purpose adsorbent — it's used to strip fluoride out of water and soil. In a paste, it can work against the other ingredients in the tube.
Where abrasion actually costs you
If you take one mechanism from this article, make it this one, because it explains a complaint I heard constantly and never had a good answer for.
Cumulative hard tissue loss exposes dentin. Dentin exposure is a well-documented route to hypersensitivity. The chain runs: wear, then recession or thinning, then exposed dentin, then the cold-water reaction.
The sensitivity does not arrive at the same time as the wear. It arrives later — often much later — which is exactly why almost nobody connects the two. By the time your teeth start reacting to cold, the material that would have insulated them has been gone for years.
And this is where the biology becomes unforgiving.
Enamel is the only tissue in the human body that cannot heal itself.
Cut your skin and it closes. Break a bone and it remodels — bone is living tissue, with cells and a blood supply and an active repair cascade. Enamel has none of that. It is roughly 96% mineral by weight, acellular, with no blood supply and no cells to rebuild with. Once formed, the only repair available to it is chemical: mineral ions redepositing onto the crystal structure that survived.
That single fact is why remineralisation products exist as a category at all. The tooth cannot fix itself biologically, so the chemistry of the environment it sits in is the only lever anyone has.
Removal is not repair
Once you see the category through this lens it's difficult to unsee.
Nearly every product in the whitening and natural-cleaning aisle is subtractive. Charcoal, silica, polishing agents, peroxide — different mechanisms, same direction of travel. All of them take something off the tooth. Some do it gently and some don't, but none of them add anything back.
Against a tissue with no capacity for self-repair, that is a finite budget being spent twice a day.
Hydroxyapatite is the only common oral care ingredient that runs the other way.
Hydroxyapatite is what enamel is made of. Delivered as nanoscale particles, it deposits into the micro-porosity of the enamel surface — the same mineral, arriving as finished material rather than being extracted from what's left. It is not a coating and it is not a polish. It is the substrate itself.
A randomised crossover clinical trial published in BDJ Open in 2019 compared 10% hydroxyapatite against fluoride toothpaste and found comparable remineralisation of early enamel lesions.
Comparable results, opposite direction. One works by adding.
Why concentration is the whole game
Here's the catch, and it's why a lot of hydroxyapatite products underdeliver.
The clinical work demonstrating meaningful remineralisation uses concentrations around 10%. A large share of consumer products entering the category sit at 2–5% — enough to justify the word on the label and the price on the shelf, not enough to match the trials.
After looking at what was actually available, one product was formulated at the concentration the research uses.
10% nano-hydroxyapatite. The concentration used in the clinical literature, not a trace inclusion — where most of the category sits at two to five per cent.
Low abrasion. Formulated to clean without spending enamel to do it.
No SLS. The detergent associated with mucosal irritation and canker sores in susceptible people.
Fluoride-free. For people who want it that way, without giving up remineralisation to get there.
90-day money-back guarantee. Three months is long enough to cover a full checkup cycle. If it isn't working, every dollar back.
| Herblix — 10% nHAP | Abrasive whitening paste | |
|---|---|---|
| Direction | Adds mineral | Removes material |
| Mechanism | Deposits into enamel porosity | Physical stain abrasion |
| Concentration disclosed | 10%, stated | Abrasivity rarely disclosed |
| Sensitivity | May help reduce over time | Cumulative wear can worsen |
| Stain result | Gradual | Faster, at a material cost |
Neither approach is magic. Abrasion genuinely does remove surface stain, and it does it faster. The question this article is asking is what it costs, and whether you were told the price.
Try Herblix risk-free →🛡 90-day money-back · 🚚 Free shipping · 🔒 Secure checkout
What to expect
Mineral deposition is not fast. It's the same slow chemistry that maintains enamel in the first place, running against decay instead of behind it.
- Week 1
- Nothing dramatic. What stops is the daily removal.
- Weeks 2–4
- Surfaces may feel smoother. Some people notice reduced sensitivity to cold in this window, though it varies considerably.
- Weeks 5–8
- The timeframe at which remineralisation effects are typically measured in the clinical literature.
- ~90 days
- A full checkup cycle — long enough to have an informed conversation with your dentist about what changed.
Questions people ask
Is my toothpaste too abrasive?
Most likely it's within the ISO limit, since products above 250 have largely been removed from the market. But "within the limit" covers everything from 4 to 250, and the limit is a safety ceiling rather than a recommendation. If you have gum recession, exposed root surfaces, or existing sensitivity, the lower end of the range matters considerably more for you than for someone with intact enamel.
Is baking soda bad for my teeth?
No, and this is the most common misconception in the whole area. Plain baking soda scores 7 — one of the gentlest abrasives in oral care, with hardness roughly equal to dentin and lower than enamel. JADA has noted it may actually suit people at high risk of erosion. Counterintuitively, toothpastes with higher baking soda content tend to score lower, because the bicarbonate is displacing harsher abrasives. If you've been avoiding it on abrasion grounds, you were misinformed.
So is charcoal actually harmful?
The honest answer is that nobody has established either way, and that's the concern. Tested charcoal pastes span a wide abrasivity range that overlaps ordinary toothpaste. What the ADA's literature review found was insufficient data to substantiate the safety and efficacy claims being made for them — and no charcoal product has earned the ADA Seal. It's an unvalidated category being sold on validated-sounding claims.
Why isn't the abrasivity number on the packaging?
There's no requirement that it be. Products must fall below the ISO ceiling; there's no obligation to disclose where below it they land. Some manufacturers publish on request. Most don't.
What if it doesn't work for me?
90-day money-back guarantee. Three months, a full checkup cycle, every dollar back if it isn't doing anything for you.
The number was always there
I'm not trying to talk you out of your toothpaste. Plenty of high-abrasivity products are perfectly reasonable for people with healthy enamel and no recession, and gentle isn't automatically better if a product doesn't clean.
What bothers me — what bothered me in that room, watching a machine measure exactly how much of a tooth was gone — is that the measurement exists, has existed for seventy years, is run routinely, and simply isn't shared. You are the only person in the transaction who doesn't get the number.
And the tissue it applies to is the one your body cannot rebuild.
Removal isn't repair. The only question worth asking about anything you brush with twice a day is which direction it's moving.
Try Herblix risk-free →🛡 90-day money-back · 🚚 Free shipping · 🔒 Secure checkout