Everything else on this site is arithmetic. This is the one place where you can replace an assumption with a measurement — and on a car calibrated for a specific blend, it is the single highest-value thing you can do.
Why the pump label is not enough
ASTM D5798 permits E85 to contain anywhere from 51% to 83% ethanol. That is not a tolerance band around 85 — it is the actual legal range, and the top of it is below the number on the sign. Content varies by season, by region, by supplier, and sometimes between two stations on the same road.
The water separation test
The standard field method relies on a simple property: ethanol is miscible with water, gasoline is not. Add a known volume of water to a known volume of fuel, shake, and let it settle. The ethanol leaves the fuel and joins the water, so the water layer grows by exactly the volume of ethanol that was in the sample.
In practice you use a graduated cylinder or a purpose-made tester. Fill to a marked line with water, add fuel to a second marked line, cap it, invert it several times, and let it stand until the layers separate cleanly. The rise in the water layer, read against the graduations, is your ethanol percentage.
Getting a reading you can trust
- Let it settle properly. A hazy interface read too early is the most common source of error. Give it time to go clear.
- Read at eye level, at the meniscus. Parallax across a wide cylinder is worth a percentage point or two.
- Take the sample from the pump you will actually use, not a different pump at the same station.
- Test near when you fill. A reading from three weeks ago describes a delivery that has probably been replaced.
What the test does not tell you
It gives you ethanol content and nothing else. It does not tell you the octane of the gasoline fraction, whether the fuel is contaminated, or whether it is fresh. It also assumes the fuel had no dissolved water in it to begin with — normally a safe assumption, and not one worth worrying about for blending purposes.
Electronic sensors
A flex-fuel content sensor plumbed into the fuel system reads continuously and feeds the ECU directly, which is a considerably better answer than testing by hand — if the vehicle and the calibration support one. For a car without that hardware, the bottle test is the practical option, and it is accurate enough for blending.
What to do with the number
Enter it on the calculator under "I tested it myself." Your result stops being a range and becomes a single number, because the uncertainty you were carrying has been measured away. That is the entire point.
Palm Beach Dyno sells a branded ethanol content test kit through Lethal Performance. Get in touch about testing and tuning →
Written by the Palm Beach Dyno team — a Ford performance shop running dyno and remote calibration on gasoline, ethanol blends and E85. Last reviewed 4 August 2026.