The Scoville Scale Explained: Where the Numbers Come From

Wilbur Scoville was a pharmacist working for Parke-Davis in Detroit, and in 1912 he was trying to solve a dull problem: how to standardise the amount of capsicum in a muscle rub. His answer was a tasting panel, and it accidentally became the most recognised food measurement in the world.

The original test used people

Scoville dissolved a measured weight of dried pepper in alcohol to extract the capsaicinoids, then diluted that extract in sugar water. He gave it to a panel of five tasters. If three of them detected heat, he diluted it further and tried again. The dilution at which the heat disappeared was the score.

So a figure of 1,000 SHU meant one part extract to a thousand parts sugar water. A figure of 1,000,000 meant a million-fold dilution before five people could no longer tell. That is why the numbers get so large so quickly, and it is a genuinely intuitive thing to measure — the score answers "how far can this be spread before it stops mattering", which is closer to how heat behaves in a dish than any laboratory value would be.

It was also, obviously, subjective. Tasters fatigue. Five people are not a population. The panel's sensitivity drifted across a session, which biased later samples downward.

What replaced it

Since the 1980s, figures come from high-performance liquid chromatography. HPLC separates a sample into its components and measures the concentration of the individual capsaicinoids — mostly capsaicin and dihydrocapsaicin, which together account for the great majority of the heat — in parts per million.

That result is called an ASTA pungency unit, and it is multiplied by roughly 16 to turn it back into a Scoville figure. There is no scientific reason for that final step. It exists because "1,641,183 Scoville" means something to the public and "102,574 ppm capsaicinoids" does not. The world's hottest-pepper records are set in Scoville units that were never tasted by anyone.

Why the same pepper measures differently twice

Capsaicin is a defence compound, and a plant under stress makes more of it. Heat within a single variety varies several-fold with sunlight, water availability and ripeness — the same jalapeño cultivar can produce pods three times hotter in a dry season than a wet one. Position on the plant matters. So does how long the pod stayed on it.

This is why every honest published figure is a range, and why the averages on the scale should be read as the middle of a wide spread rather than a specification. It is also the reason a training programme that trusted chart numbers would be building on sand: what matters is the burn you actually report, not the number the pepper was supposed to be.

Where the heat lives

Not in the seeds, which is the most durable myth about peppers. Capsaicin is produced in glands on the placenta — the pale pith the seeds attach to. Seeds pick up heat by contact only. Removing the seeds works because you take the pith out with them.

Why linear numbers need a logarithmic ladder

The Scoville numbers themselves are perfectly linear: 100,000 really is twice 50,000. Perception is not. Each doubling of concentration produces about the same increase in perceived intensity wherever on the scale it happens, which is the same relationship that governs loudness and brightness.

The practical consequence is severe. Plot every pepper in the world on a straight line from zero to Pepper X at 2,693,000 SHU and everything most people have ever eaten — every jalapeño, every bottle of Sriracha, all of Cayenne — collapses into the first two percent of the distance. A step that is huge at the bottom of that line is invisible at the top.

Any ranking meant to be climbed has to be spaced in multiples instead. That is what a ladder is for, and it is why the app has twelve rungs rather than a number you are supposed to interpret.