Scoville Chart

How the Scoville scale works

The Scoville scale gives chillies a number for their heat. Here is where the number comes from, why it is only ever "about", and what the scale cannot tell you.

The original test

In 1912 the American pharmacist Wilbur Scoville devised a way to rate how hot a chilli is (Scoville, 1912, cited in Bosland and Walker, 2010). A sample of the chilli was soaked in alcohol to pull out the compounds that cause the burn. That extract was then diluted in sugar water, a little more each time, and a panel of tasters sipped it. The test stopped when the tasters could no longer detect any heat. Germany's Federal Institute for Risk Assessment describes the panel as five people, with at least three having to agree that the heat was gone (BfR, 2024). The alcohol, the sugar water and the tasting panel are also described by Birchall (2018).

The number of times the extract had to be diluted is the Scoville rating, given in Scoville heat units, or SHU. In effect, the rating is the dilution factor. You can see what that dilution looks like for any chilli on the chart. If the extract had to be diluted 5,000-fold before the heat vanished, the chilli was rated at 5,000 SHU. A bell pepper needs no dilution at all, so it rates zero (Zepp (2026) notes the scale starts at 0, the heat of a bell pepper).

Scoville described his test in a short paper, "Note on capsicums", published in 1912. We have not read it ourselves, so we cite it as described by Bosland and Walker (2010), which is also a plain-language guide to how chilli heat is measured. A 1920 textbook, quoted by DeWitt (2012), gives Scoville's recipe as soaking ground chilli overnight in alcohol and adding the strained liquid to sweetened water (10 per cent sugar) until a weak heat was just noticeable (Lyons, 1920, cited in DeWitt, 2012). Wikipedia (no date a) supports only that he was a pharmacist working for Parke-Davis and that his 1912 test used human testers.

How it is measured now

Tasting is subjective, since it depends on each taster's palate and sensitivity, and a taster can only handle so many samples. So laboratories now usually use a machine method called high-performance liquid chromatography (HPLC) (Koester, no date; Bosland and Walker, 2010). It measures the amount of capsaicin, said kap-SAY-ih-sin (Cambridge Dictionary, no date), and related compounds, together called capsaicinoids, in the sample.

The machine result is then converted to Scoville units. The conversion factor is not fully settled: published sources use figures from 15 to about 16 SHU for each part per million of capsaicinoid. Germany's Federal Institute for Risk Assessment states the convention as 16 SHU for each milligram of capsaicin per kilogram, which is the same thing (BfR, 2024). A 2026 systematic review by Hayward et al. (2026) notes a factor of 16.1 from a 1977 study and others between 15 and 16.1, and calls this an ongoing inconsistency. New Mexico State University's own 2012 study of superhot chillies used a factor of 16 (Bosland et al., 2012). On a factor of 16, pure capsaicin comes out at around 16 million SHU, which matches the figure usually quoted for the top of the scale. Because laboratories can use different factors, published figures for the same chilli do not always match.

Why every figure is a range

You will rarely see one exact number for a pepper here, and that is deliberate. Heat varies with:

New Mexico State University's chilli researchers describe heat as the result of a plant's genetics together with how it interacts with its environment. They say a mild variety will become hotter when it is stressed in the field, a hot variety grown in ideal conditions may be only moderately hot, and a plant bred for low heat will not become truly hot even when stressed (Bosland and Walker, 2010). The same university's 2012 trial of superhot chillies shows how wide the spread can be: single ghost pepper plants ranged from about 279,000 to about 1.58 million SHU, although the average was about 1.02 million (Bosland et al., 2012). A record-setting figure comes from the samples that were tested, so it does not mean every pod of that variety will match it.

What the scale does not tell you

How "times hotter" is worked out here

The compare tool takes the middle of each published range and divides one by the other. It also shows the gap you would get if you compared the top of one range with the bottom of the other, so you can see how wide the uncertainty is. If one item scores zero, no ratio is shown, because dividing by zero is meaningless. If two ranges overlap, the tool says so.

Our heat bands

The Scoville scale has no official named levels. The bands on the chart are our own grouping, chosen to make the chart easier to read:

BandMiddle of range, in SHU
MildUnder 2,500
Medium2,500 to 25,000
Hot25,000 to 100,000
Very hot100,000 to 500,000
Superhot500,000 to 1,500,000
ExtremeOver 1,500,000

Handling chillies safely

This is general guidance, not medical advice.

References

Back to the chart