Problems · brakes

Old DOT 4 boils lower than new DOT 3.

Brake fluid is one of the few things on a car whose performance is set by federal law rather than by a marketing department. The law gives every grade two numbers, and the gap between them is the whole story: absorbing moisture costs about a third of the boiling point, whichever grade you bought.

Moisture costs about a third of it

Glycol brake fluid is hygroscopic — it pulls water out of the air, through seals and hoses, over time rather than over distance. The regulation accounts for this by setting two minimums for every grade: one for fresh fluid, and one after a specified humidification procedure.

Every fresh fluid beats every wet one

Put all six states in one column, ordered by the temperature that actually protects you, and the grades stop being the story:

The line falls between fresh and wet, not between grades. Fresh DOT 3 — the cheapest fluid on the shelf — boils 50 °C higher than wet DOT 4 and 25 °C higher than wet DOT 5. There is no grade you can buy that outruns neglect.

Which makes the practical rule short. Use the grade your manual specifies — and never step down a grade, since a system designed for DOT 4 is designed around DOT 4’s numbers. Then change it on the schedule in the manual, because the clock and not the odometer is what is spending it.

The law puts the number on the bottle

The regulation requires the container to carry the minimum wet boiling point, in Fahrenheit. Not the dry one. The number that describes the fluid after it has been in your car is the number the law insists is printed, and almost nobody reads it.

That is what makes this category comparable without any test at all. Two bottles of the same grade are both legal at the same floor, but the one printing a higher wet figure is committing to more margin, in a unit you can compare on the shelf. It is the same shape as a torque wrench’s certified tolerance or the MIN TH cast into a rotor: a standard is what turns a number into a claim you can hold someone to.

Two more label facts from the same section, both useful in the aisle: DOT 3, DOT 4 and DOT 5.1 non-silicone fluids must be colourless to amber, while DOT 5 silicone must be purple — and a DOT 5 container has to say whether it is “silicone base” or “non-silicone base”, because DOT 5 and DOT 5.1 are not the same fluid despite the names.

What boiling actually does to the pedal

A boiling point matters because a hydraulic system works only on liquid. Liquid does not compress; vapour does. When hard or repeated braking heats the fluid past its boiling point, bubbles form in the lines and the pedal now compresses those bubbles instead of pushing the pistons.

The symptom is a pedal that goes long or straight to the floor with the brakes still nominally working, and it arrives without warning on the descent where you needed them. Brembo describe this failure and put the onset at a fluid temperature above 200 °C.

Put that threshold on the table above and it lands in the gap. 200 °C is higher than the wet minimum of every grade — 180, 155 and 140 °C — and lower than the fresh minimum of every grade — 205, 230 and 260. Fresh fluid of any grade is legally required to survive it. Fluid that has taken up moisture, of any grade, is not.

What to buy, and what we won't recommend

Fluid is not vehicle-specific the way pads and tyres are — it is grade-specific — so you can genuinely shop it. What you cannot do is rank brands from anything published, so we are not going to pretend otherwise. Buy the grade in your manual and compare the printed wet boiling points, which is a real comparison the law hands you for free.

• DOT 4 brake fluid on Amazon Compare the wet boiling point printed on each bottle — legal minimum 311 °F
• DOT 3 brake fluid on Amazon Only if your manual specifies DOT 3 — legal minimum 284 °F wet

Affiliate links — we earn a commission at no extra cost to you and it does not change what we recommend. Brake fluid absorbs moisture from an opened bottle too; buy the size you will use. How this site makes money.

We are not recommending a brake fluid tester. The cheap LED pens sold for this read electrical conductivity, which is a proxy for water content and not a boiling-point measurement, and we could find no independent validation of how well the proxy holds. Where we cannot find a measurement we do not rank the products — the same position we take on jack stands.

What nobody has measured

1. How fast fluid actually absorbs water in your car. The regulation defines a laboratory humidification procedure, not a calendar. How long a real system takes to reach the wet figure depends on climate, seals and how often the reservoir is opened, and no published curve covers it.
2. Which brand holds its margin longest. Every product must clear its grade's floor; nothing published compares how far above it they start or how quickly they fall.
3. What the pens are worth. No independent test we could find checks a conductivity reading against an actual boiling point on the same sample. Until one does, the reading is an indication and not a measurement.