What an IBU is, and what it is not
An International Bitterness Unit is a laboratory measurement: one milligram per litre of isomerised alpha acid, determined by extracting a beer sample into iso-octane and reading its absorbance at 275 nm. It is a chemical concentration, not a taste score. A recipe IBU number, including the one on this page, is a prediction of that measurement, and the gap between the two is where most of the confusion about bitterness lives.
Hops contribute bitterness through their alpha acids, which are almost insoluble in wort as they come off the plant. Boiling rearranges them into iso-alpha acids, which are soluble and intensely bitter. The reaction is slow, which is why bittering hops go in for an hour, and it is incomplete, which is why utilization is a number well below one. Anything that suppresses isomerisation or drops iso-alpha acid out of solution before packaging — a dense sugary wort, a vigorous fermentation carrying compounds out on the foam, a hop-heavy trub bed left behind in the kettle — pushes the measured figure below the calculated one.
So treat the number as a recipe currency. It lets you scale a recipe to a new batch size, substitute one hop for another, and compare your beer against style guidelines. It does not tell you how bitter the beer will taste, because perceived bitterness depends just as much on the residual sweetness, the water sulphate-to-chloride balance, the carbonation and the alcohol.
The Tinseth model, factor by factor
Glenn Tinseth built his utilization model in the 1990s from measured bitterness in real batches, and its shape reflects two observations every brewer can confirm. Utilization rises with boil time but flattens out; and utilization falls as the wort gets denser.
The boil time factor is (1 − e−0.04t) / 4.15. That is a saturating exponential: it climbs quickly through the first twenty minutes, is well past halfway by thirty, and is nearly flat by ninety. At 1.050 boil gravity, going from 15 to 30 minutes takes utilization from 11.4% to 17.7% — a big gain. Going from 60 to 90 minutes takes it from 23.1% to 24.7%, a gain so small that the extra half-hour of boiling costs you more in evaporation and colour than it buys in bitterness. This is the single most useful thing the model tells you.
The bigness factor is 1.65 × 0.000125(Gb−1), where Gb is the average gravity of the wort during the boil. It is an exponential decay in gravity points. At 1.030 it is 1.26; at 1.050 it is 1.05; at 1.080 it is 0.80. In practical terms, the same hop charge boiled in a strong wort delivers roughly a third less bitterness than in a weak one. Two consequences follow. A high-gravity beer needs a disproportionate weight of hops to reach the same IBU, and a concentrated partial boil topped up with water afterwards will under-bitter unless you use the concentrated wort's gravity in the calculation, not the finished beer's.
The volume divisor is post-boil volume, because IBU is a concentration. Boiling off harder does not create bitterness, but it does concentrate it into fewer litres, which raises the IBU of the finished beer.
Multiply the three and you have utilization; multiply that by the milligrams of alpha acid in the addition and divide by litres, and you have IBUs. Every addition is calculated independently and the results add.
Worked example: an American IPA at 5 US gallons
Five US gallons (18.927 L) post-boil, average boil gravity 1.050, original gravity 1.062, whole hops. Three additions: 1 oz (28.35 g) of 12% alpha Magnum at 60 minutes, 1 oz of 5.5% alpha Cascade at 15 minutes, and 1 oz of the same Cascade at 5 minutes.
- Bigness factor. 1.65 × 0.0001250.050. The exponent gives 0.6380, so the factor is 1.0527.
- 60-minute time factor. (1 − e−2.4) ÷ 4.15 = (1 − 0.09072) ÷ 4.15 = 0.21910.
- Utilization at 60 minutes. 1.0527 × 0.21910 = 0.23066, or 23.07%.
- IBU from the Magnum. 0.12 × 28.35 × 1000 × 0.23066 ÷ 18.927 = 41.5 IBU.
- 15-minute utilization. Time factor (1 − e−0.6) ÷ 4.15 = 0.10872; times 1.0527 gives 0.11445, or 11.45%.
- IBU from the 15-minute Cascade. 0.055 × 28.35 × 1000 × 0.11445 ÷ 18.927 = 9.4 IBU.
- 5-minute utilization. (1 − e−0.2) ÷ 4.15 = 0.04368; times 1.0527 gives 0.04598, or 4.60%.
- IBU from the 5-minute Cascade. 0.055 × 28.35 × 1000 × 0.04598 ÷ 18.927 = 3.8 IBU.
- Total. 41.5 + 9.4 + 3.8 = 54.7 IBU.
- BU:GU. 54.7 ÷ 62 gravity points = 0.88.
Note how the arithmetic settles a common argument: the two late Cascade additions together, two-thirds of the hop weight, supply less than a quarter of the bitterness. They are there for aroma, and the model is telling you so.
Reading total IBU and the BU:GU ratio
Absolute IBU only makes sense next to gravity, which is why the BU:GU ratio matters more than the raw number. Divide total IBU by the original gravity points — 54.7 IBU on a 1.062 wort is 54.7 ÷ 62 = 0.88. Around 0.5 the beer reads malt-forward; near 0.75 it reads balanced; above 1.0 it reads firmly bitter. A 40 IBU bitter at 1.040 (ratio 1.00) tastes far more bitter than a 40 IBU doppelbock at 1.075 (ratio 0.53), and that is exactly what the ratio predicts.
Compare your total against the style you are aiming at: most lagers and wheat beers sit between 8 and 25 IBU, pale ales between 30 and 50, American IPAs between 40 and 70, and imperial stouts and double IPAs from 60 upwards. Judges score against these ranges, so a recipe that misses by twenty points reads as the wrong beer even when it is well made.
Above roughly 100 calculated IBU the number stops being physically meaningful. Wort cannot hold much more than that in solution, so a recipe calculating to 150 will measure far lower in the glass. Recipes advertising three-figure bitterness are quoting arithmetic, not analysis.
Use the grams-per-IBU output to hit a target rather than guessing. It is the weight of your first hop, at its alpha acid and boil time, that supplies exactly one IBU. Want 15 more IBU from that addition? Multiply. Substituting one hop for another at the same time is even simpler: the bitterness is proportional to weight times alpha acid, so 1 oz at 12% is replaced by 2 oz at 6%. Check the resulting gravity and efficiency with the brewhouse efficiency calculator so your boil gravity input matches the wort you actually make.
Tinseth utilization (%) by boil time and boil gravity
| Boil gravity | 10 min | 15 min | 20 min | 30 min | 45 min | 60 min | 75 min | 90 min |
|---|---|---|---|---|---|---|---|---|
| 1.030 | 10.0 | 13.7 | 16.7 | 21.2 | 25.3 | 27.6 | 28.9 | 29.5 |
| 1.040 | 9.1 | 12.5 | 15.3 | 19.4 | 23.2 | 25.2 | 26.4 | 27.0 |
| 1.050 | 8.4 | 11.4 | 14.0 | 17.7 | 21.2 | 23.1 | 24.1 | 24.7 |
| 1.060 | 7.6 | 10.5 | 12.8 | 16.2 | 19.4 | 21.1 | 22.0 | 22.6 |
| 1.070 | 7.0 | 9.6 | 11.7 | 14.8 | 17.7 | 19.3 | 20.1 | 20.6 |
| 1.080 | 6.4 | 8.7 | 10.7 | 13.5 | 16.2 | 17.6 | 18.4 | 18.8 |
Read across a row to see how little the last half-hour buys: at 1.050 the jump from 60 to 90 minutes is 1.6 percentage points, against 6.3 points for the jump from 15 to 30 minutes. Read down a column to see the gravity penalty: the same 60-minute charge in 1.080 wort delivers about three-quarters of the bitterness it would in 1.040 wort.
Tinseth, Rager and Garetz give different answers on purpose
Three utilization models circulate in homebrewing and they are not interchangeable. Tinseth, used here, generally returns the lowest figures of the three for late additions and is the default in most modern software. Rager's model returns noticeably higher numbers, particularly at short boil times, and includes a gravity adjustment that only kicks in above 1.050. Garetz adds corrections for hopping rate, elevation and yeast handling and is the most conservative. None of them is wrong — each was fitted to a particular brewery's measurements — but a recipe published with a Rager IBU and rebuilt with Tinseth will look under-bittered. Always note which model a recipe used before you match its numbers.
Where an IBU estimate goes astray
- Using the finished beer's gravity as the boil gravity. On a partial boil the wort in the kettle is much denser than the topped-up batch, so utilization is lower than you think and the beer comes out under-bittered.
- Trusting a variety's alpha acid instead of the packet's. Cascade can be lot-tested anywhere from 4.5% to 8.9%. Using a book figure instead of the number printed on your bag is routinely worth ten IBU on a bittering charge.
- Ignoring hop age and storage. Alpha acids oxidise. Hops kept warm and in air lose a large share of their bitterness potential within a year; vacuum-packed and frozen, they hold it for years.
- Counting whirlpool and dry hops as boil additions. They contribute aroma and some measurable bitterness, but not by the boil-time mechanism this model describes. Enter them as zero minutes and treat any extra bitterness as a bonus.
- Expecting the measured IBU to match. Fermentation, trub loss and filtration all remove iso-alpha acid. Beers that are heavily dry-hopped or fermented under pressure often measure well below the recipe figure.
- Forgetting that boiling harder raises IBU. Bitterness is a concentration. Losing an extra half-gallon to evaporation concentrates the same iso-alpha acids into less beer; work your volumes out first with the mash and sparge water calculator.
Where bitterness fits in the rest of the recipe
Bitterness is one leg of a three-legged balance, and the other two come from calculations you should do at the same time. Gravity sets the malt sweetness the bitterness has to cut through, so a recipe is only balanced once you know your real efficiency and therefore your real original gravity. Attenuation sets how much of that sweetness survives fermentation: the same 50 IBU beer tastes sharply bitter at 78% apparent attenuation and rounded at 68%, which you can check afterwards with the ABV and attenuation calculator.
Mash temperature is the lever for the third leg. A mash held at 148 °F makes a highly fermentable wort and a dry, bitter-leaning beer; the same grain bill at 156 °F leaves more dextrin and pads the bitterness. Set that deliberately with the strike water temperature calculator rather than discovering it at the end.
Finally, remember what the model cannot see. Sulphate in the brewing water sharpens bitterness perceptibly; chloride softens it. Carbonation level changes the attack: the same beer at 2.8 volumes reads more bitter than at 2.0, which is worth thinking about before you reach for the priming sugar calculator. And polyphenols from a large late-hop charge add a drying astringency that no IBU figure captures. The number is a starting point that gets you within a few points of the beer you intended; your palate closes the gap.