Two families of unit that look interchangeable and are not, a Saybolt scale that is a curve rather than a factor, and oil grades that are ranges rather than numbers.
Calculator
Try:
100 centipoise (cP)
100.0 cP
A viscosity without a temperature is not a fact — most liquids change by an order of magnitude across an ordinary working range.
Dynamic units
The dynamic family — force units, no density involved
Unit
Value
What it is
pascal-second (Pa·s)
0.10000
The SI unit. Equal to one newton-second per square metre.
millipascal-second (mPa·s)
100.0
Identical to the centipoise, which is why both appear on the same data sheets.
poise (P)
1.000
The CGS unit, named for Poiseuille. Exactly a tenth of a pascal-second.
centipoise (cP)
100.0
A hundredth of a poise, and the unit most industrial figures are quoted in. Water is about 1 cP at 20 °C.
lbf·s/ft²
0.00209
The US customary unit, also called the slug per foot-second.
lb/(ft·s)
0.06720
Pound-mass per foot-second — the absolute imperial form, 32.174 times smaller than lbf·s/ft².
The other family is deliberately not in this list. Centipoise and centistokes are different quantities, and the only thing that joins them is the density — which is the panel below, not a row in this table.
Crossing between them needs a density
Kinematic viscosity113.6 cStν = μ ÷ ρ
The same 100.0 cP, in fluids of different density
Fluid
Density
Kinematic viscosity
Water
0.9982 g/cm³
100.2 cSt
Petrol / gasoline
0.72 g/cm³
138.9 cSt
Diesel
0.83 g/cm³
120.5 cSt
SAE 30 engine oil
0.88 g/cm³
113.6 cSt
Olive oil
0.91 g/cm³
109.9 cSt
Honey
1.42 g/cm³
70.423 cSt
Glycerol
1.26 g/cm³
79.365 cSt
Mercury
13.534 g/cm³
7.389 cSt
Water is about 1 in both units at 20 °C, and that coincidence is the whole reason the two get treated as interchangeable. Move to mercury — thin but dense — and the same dynamic viscosity is eighteen times smaller in kinematic terms.
Saybolt seconds, and why 4.63 is not a factor
Saybolt Universal Seconds526.6ratio 4.634 to the cSt
The Saybolt ratio at different viscosities
Kinematic viscosity
Saybolt seconds
SUS per cSt
2 cSt
32.6
16.301
5 cSt
42.4
8.479
10 cSt
58.8
5.884
32 cSt
150.5
4.703
100 cSt
463.5
4.635
460 cSt
2,130.9
4.632
1000 cSt
4,632.4
4.632
The ratio runs from 16.3 at 2 cSt down to about 4.632 at high viscosity — a factor of 3.5 in a number that conversion charts routinely print as a constant. It is a formula, not a multiplier, and it has no closed-form inverse; the reverse direction here is solved numerically rather than by rearranging an approximation.
What SAE grade that is, at 100 °C
Grades satisfiednoneif this figure is at 100 °C
SAE J300 engine oil grades, measured at 100 °C
Grade
Range (cSt)
Spread
SAE 8
4 – 6.1
52%
SAE 12
5 – 7.1
42%
SAE 16
6.1 – 8.2
34%
SAE 20
6.9 – 9.3
35%
SAE 30
9.3 – 12.5
34%
SAE 40
12.5 – 16.3
30%
SAE 50
16.3 – 21.9
34%
SAE 60
21.9 – 26.1
19%
The W half is a different test at a different temperature
Winter grade
Cranking limit
Tested at
0W
6,200 cP max
-35 °C
5W
6,600 cP max
-30 °C
10W
7,000 cP max
-25 °C
15W
7,000 cP max
-20 °C
20W
9,500 cP max
-15 °C
A grade is a range, not a value: SAE 30 is anything from 9.3 to 12.5 cSt, so two compliant 30-weight oils can differ by a third. And the two halves of “5W-30” are not two points on one scale — the 30 is kinematic viscosity in centistokes at 100 °C, while the 5W is a dynamic cranking limit in centipoise at −30 °C. A 5W-30 and a 10W-30 have identical hot requirements.
What this converter covers
Dynamic and kinematic units listed separately, bridged only through a density you supply — with Saybolt seconds and the SAE grade a figure actually meets.
Pascal-seconds, poise, centipoise and the two imperial dynamic units
Square metres per second, stokes, centistokes and ft²/s
The density bridge between the two families, in both directions
Saybolt Universal Seconds, forward and numerically inverted
SAE J300 grades at 100 °C and the W cranking limits
Two families, kept apart Density required Saybolt is a curve Temperature stated
100 cP is 100 cSt in water and 7.4 in mercury · SAE 30 is anything from 9.3 to 12.5 cSt
Centipoise and centistokes are not the same thing
Two liquids can have the same centipoise figure and behave completely differently, because there are two viscosities and only one of them is what people usually mean.
Dynamic viscosity (μ, in centipoise) is resistance to shear — how much force it takes to slide one layer of fluid over another. Kinematic viscosity (ν, in centistokes) is that divided by density: how the fluid flows when the thing driving it is its own weight. A viscometer that measures how long a sample takes to drain through an orifice is measuring kinematic viscosity, because gravity is doing the pushing.
ν = μ/ρ. So converting between them requires the density, and there is no universal factor. This tool asks for one and refuses without it, which is occasionally annoying and always correct.
The same 100 cP, in fluids of different density
Fluid
Density (g/cm³)
Kinematic viscosity
Petrol
0.72
138.9 cSt
Engine oil
0.88
113.6 cSt
Water
1.00
100.0 cSt
Honey
1.42
70.4 cSt
Mercury
13.53
7.4 cSt
Nineteen-fold, from one dynamic viscosity. The reason people miss it is a coincidence: water at 20 °C is about 1 cP and about 1 cSt, because its density is about 1 g/cm³. Every intuition built on water quietly assumes a density of one, and every fluid that is not water breaks it.
Mercury is the case that makes the distinction obvious. It is not viscous — it flows more freely than water — but it is very dense, so its kinematic viscosity is tiny while its dynamic viscosity is comparable to water’s. Which of those matters depends entirely on what is pushing the fluid.
Saybolt seconds are not a multiple
Saybolt Universal Seconds is an older American scale: the time in seconds for 60 ml of a sample to drain through a standard orifice. It survives on data sheets and in specifications, and conversion charts routinely give the factor as 4.63.
That number is only right at the top of the range. The relation is a curve, and the ratio between Saybolt seconds and centistokes falls as viscosity rises.
The Saybolt ratio is not constant
Kinematic viscosity
Saybolt seconds
SUS per cSt
2 cSt
32.6
16.30
5 cSt
42.4
8.48
10 cSt
58.8
5.88
32 cSt
150.6
4.71
100 cSt
463.5
4.63
1000 cSt
4632.4
4.63
A factor of 3.5 from one end to the other. Using 4.63 on a light hydraulic fluid at 2 cSt gives 9 seconds where the answer is 33 — not a rounding error but a different number entirely.
The shape has a physical cause. At high viscosity the drain time is dominated by the fluid’s resistance and scales linearly with it. At low viscosity the fluid’s own momentum matters — a kinetic-energy correction — and it drains slower than proportionality predicts. Below about 1.8 cSt the scale stops resolving usefully, which is why this tool returns nothing there rather than extrapolating.
One consequence worth knowing: the relation has no closed-form inverse. Going from Saybolt seconds back to centistokes cannot be done by rearranging the formula, so this page solves it numerically. Charts that offer a reverse factor have quietly linearised something that is not linear.
An SAE grade is a range
“SAE 30” is not a viscosity. It is a band, and any oil inside it may be sold as 30 weight.
SAE J300 grades at 100 °C, with the spread inside each
Grade
Range (cSt)
Spread
SAE 20
6.9 – 9.3
35%
SAE 30
9.3 – 12.5
34%
SAE 40
12.5 – 16.3
30%
SAE 50
16.3 – 21.9
34%
SAE 60
21.9 – 26.1
19%
Two compliant 30-weight oils can differ by a third. Both meet the standard; neither is mislabelled. If a specification calls for a viscosity rather than a grade — and engine and gearbox manufacturers increasingly do — the grade alone does not satisfy it.
The newer light grades add a second surprise: they overlap. SAE 12 runs 5.0–7.1, SAE 16 runs 6.1–8.2 and SAE 20 runs 6.9–9.3, so an oil measuring 6.95 cSt legitimately satisfies all three at once. The grades were added over time as fuel-economy formulations moved lighter, and the ladder was extended downwards rather than redrawn.
For gear oils there is a separate ladder — SAE J306, with numbers like 75W-90 — and its grades are on a different scale entirely. A gear oil marked 75W-90 is roughly comparable in viscosity to an engine oil around SAE 40, and comparing the numbers directly is meaningless. Two standards, two scales, one style of label.
5W-30 is two different tests
The two halves of a multigrade label are not two points on one scale. They are different measurements, of different quantities, in different units, at different temperatures.
What each half of “5W-30” is actually measuring
Half
Quantity
Unit
Temperature
5W
Dynamic viscosity, cranking
cP, maximum
−30 °C
30
Kinematic viscosity
cSt, a range
100 °C
So a 5W-30 and a 10W-30 have identical requirements at operating temperature. The difference between them is entirely at the cold end: how easily the engine turns over and how fast oil reaches the top of the head on a winter morning. Choosing 5W over 10W buys cold-start protection, not a different oil once warm.
This also disposes of a persistent piece of folklore — that a multigrade “gets thicker when hot”. It does not. All oils thin dramatically as they warm; a 5W-30 goes from thousands of centipoise at −30 °C to around 10 cSt at 100 °C. What a multigrade does is thin less than a straight grade would, because a viscosity-index improver added to a light base oil expands as it warms and partly offsets the thinning.
Which is why the W number is a limit rather than a range, incidentally: it is a “must not exceed” at a cold temperature, whereas the hot grade is a band the oil must land inside. Reading it as though both were the same kind of claim is where most of the confusion around multigrades starts.
Nothing here means anything without a temperature
Viscosity is more temperature-sensitive than almost any other everyday fluid property. Engine oil falls by roughly an order of magnitude between 40 °C and 100 °C. Honey at room temperature is unpourable and at 50 °C runs like syrup. Water halves in viscosity between 20 °C and 55 °C.
So “the viscosity of X is 100 cP” is not a fact. Every reference figure in this tool carries the temperature it belongs to, and every standard specifies one: SAE grades at 100 °C, ISO VG grades at 40 °C, Saybolt conventionally at 100 °F or 210 °F. When two data sheets disagree, the temperature is the first thing to check.
The rate of change is itself a specified property, called the viscosity index. A high-VI oil thins less as it warms, which is what makes it usable across a wide temperature range. It is a dimensionless number on an empirical scale rather than a physical constant, and comparing VI figures between very different base oils is less meaningful than it looks.
And one broader limit on everything above: all of it assumes the fluid is Newtonian, meaning its viscosity does not depend on how hard it is being sheared. A great many practical liquids are not. Paint, blood, ketchup, drilling mud and most greases all change viscosity with shear rate, sometimes by orders of magnitude, so a single number describes them only at a stated shear rate. Converting units for such a fluid is still valid arithmetic — but the number being converted describes one operating point rather than the material.
Related calculators
Other fluid and material property tools:
Densitykg/m³, lb/ft³, lb/in³ and lb/gal, with specific gravity against any water reference and the API scale that runs backwards.
Flow RateL/min, m³/h, GPM, CFM and barrels a day — with the two gallons, standard against actual CFM, and mass flow.
Volume to WeightGallons to pounds, cubic yards to tons, litres to kilograms — by substance, with an honest range rather than one invented number.
Pressurepsi, bar, kPa, atm, torr and mmHg — plus why gauge and absolute differ by an offset that no factor can bridge.
Fuel Economympg, L/100 km and km/L — US and UK mpg kept apart, and the reciprocal explained, since a better car has a LOWER L/100 km.
The Saybolt relation here is the ASTM D2161 approximation, which is a published curve fit rather than a derivation — this page reproduces it and does not attempt to improve on it. The SAE grade limits are the J300 classification, and the reason they are stated as ranges on this page is simply that they are ranges in the standard. ISO 3104 supplies the definition that makes the density bridge necessary and the requirement that a viscosity always carry its temperature.
SAE J300 — Engine Oil Viscosity ClassificationSAE International · verified 2026-09-07 · That each numeric grade is a RANGE of kinematic viscosity at 100 °C rather than a value, and that the W grades are separate low-temperature cranking and pumping limits stated in centipoise
These conversions are arithmetic and the fluid behaviour behind them is not. Everything here assumes a Newtonian fluid, whose viscosity does not depend on how hard it is being sheared — and many of the liquids people look up are not: paint, blood, ketchup, drilling mud, polymer solutions and most greases all change viscosity with shear rate, so a single number describes them only at a stated shear rate. Viscosity is also strongly temperature-dependent, typically by an order of magnitude across an ordinary working range, so a figure without a temperature is not usable and this page will not invent one. The reference fluid values are typical published figures at the temperature shown, not measurements of your sample. The SAE and Saybolt relations apply to petroleum products under the conditions the relevant standards specify. Where a viscosity has to be relied on — a lubrication specification, a pump selection, a process design — use the measured value on the actual product's data sheet at the actual operating temperature.
Published the Viscosity Converter: pascal-seconds, poise, centipoise and the two imperial dynamic units alongside square metres per second, stokes, centistokes and square feet per second.
Lists the dynamic and kinematic families in separate panels rather than one dropdown, because putting centipoise and centistokes in the same list is the mistake the page exists to correct.
Requires a density to cross between the families and returns nothing without one -- 100 centipoise is 100 centistokes in water, 139 in petrol and 7.4 in mercury, a nineteenfold spread from a single dynamic viscosity.
Names the coincidence that causes the confusion: water at 20 degrees is about 1 in both units because its density is about 1, so every intuition built on water quietly assumes a density of one.
Implements the ASTM D2161 Saybolt relation as a formula rather than a factor, and shows the ratio falling from 16.3 at 2 centistokes to an asymptote near 4.632 -- a factor of 3.5 across the range that conversion charts print as a constant.
Solves the reverse direction numerically, since the Saybolt relation has no closed-form inverse, and refuses below 1.8 centistokes where the scale stops resolving rather than extrapolating.
Treats an SAE grade as the range it is: SAE 30 is anything from 9.3 to 12.5 centistokes at 100 degrees, a 34 percent spread with both ends compliant, so the grade does not identify a viscosity.
Shows that the newer light grades genuinely overlap in J300, so a measurement of 6.95 centistokes satisfies SAE 12, 16 and 20 at once.
Separates the two halves of a multigrade label -- the 30 is kinematic viscosity in centistokes at 100 degrees, the 5W a dynamic cranking limit in centipoise at minus 30 -- so a 5W-30 and a 10W-30 have identical hot requirements.
Disposes of the folklore that a multigrade thickens when hot: all oils thin dramatically as they warm, and a multigrade simply thins less than a straight grade would.
States the Newtonian assumption plainly, naming paint, blood, ketchup, drilling mud and greases as fluids a single viscosity figure does not describe, and requires a temperature with every reference value.
Verified by 52 automated cases, asserting that every cross-family unit pair is refused, that the Saybolt ratio falls monotonically rather than being constant, and that the Saybolt round trip closes across four decades.
Add this calculator to your site
Responsive embed — and private: nothing your visitors type leaves their browser.