The Puree Adjustment Card: Flour, Liquid and Bake-Time Changes by Water Level
Take water out of canned pumpkin and the recipe around it stops matching. This card gives the correction for five water levels. For each one: the liquid to put back, the flour to cut instead, and how the bake time moves. All values per cup of starting purée. The full grid is below.
A 20-minute blot pulls about 39 grams of water out of one cup of canned pumpkin purée. Put those 39 grams back as liquid and the recipe behaves as written. Water level here means the share of free water you removed, weighed on a scale. Add-back means the liquid you return so the batter keeps its original ratio. There is a second route. Instead of adding liquid, cut about 43 grams of flour and keep the drier purée.
You get a smaller batch and a tighter crumb. The limit is arithmetic. Once the flour cut passes a quarter of the flour weight, it stops being a correction. It becomes a different recipe. At that point, put the liquid back instead. Bake time shifts too, but by a few minutes, never by tens.
| Level | How to get there | Purée you end up with (g) | Water removed (g, per cup of starting purée) | In tablespoons | Total solids (%) | A · Liquid to add back (g, per cup of starting purée) | B · Flour to cut instead (g, per 240 g flour) | Muffins, change from a 20 min base | 9-inch pie, change from a 55 min base |
|---|---|---|---|---|---|---|---|---|---|
| L1 | Straight from the can | 245.0 g | 0 g | 0 | 10.0 % | 0 g | 0 g | — | — |
| L2 | Paper-towel-lined bowl, one pass | 227.0 g | 18 g | 1.2 | 10.8 % | 18 g | 20 g | −0.6 min | −2 min |
| L3 | Three flat sheets, 20 min, undisturbed | 205.8 g | 39 g | 2.7 | 11.9 % | 39 g | 43 g | −1.4 min | −4 min |
| L4 | Pressed, towels renewed | 183.8 g | 61 g | 4.1 | 13.3 % | 61 g | ⚠ 67 g — use A instead | −2.2 min | −7 min |
| L5 | Reduced by half, or blotted with fresh towels until dry | 122.5 g | 123 g | 8.3 | 20.0 % | 123 g | ⚠ 133 g — use A instead | −4.8 min | −14 min |
Reference unit: 1 cup of starting purée = 245 g (USDA). Total solids retained: 24.5 g. One tablespoon of water = 14.79 g.
Read every row per cup of starting purée. Starting from a whole 15-ounce can (425 g)? Multiply by 1.735. A can taken to L3 gives you 357 g of purée and releases 68 g of water.
Column B assumes a 240 g flour base, with hydration from the purée worth 91.9 % of the flour weight. On a recipe with more flour, recalculate: flour to cut = water removed ÷ 0.919.
Past 25 % of the flour weight, column B stops being practical. Cutting 67 g of flour out of 240 g is not a correction anymore. At L4 and L5, use column A.
Every number on this card comes out of our tested guide to canned pumpkin water content. The protocols behind the second column are written up in how each water level is reached.
Where these numbers come from. Every figure on this card is calculated from published protocols and USDA composition data. The working is in the four articles it links to. The bake-time columns are model outputs, not stopwatch readings. When we have measured them, this card becomes v1.1 and says so.
How to Read the Card
The card runs on one number: the weight your purée loses. Start on a scale. Weigh the purée straight out of the can, before you do anything to it. Blot it, press it, or reduce it in a pan. Then weigh it again and subtract. That difference, divided by the starting weight, is your water level. Match your percentage to the nearest row and read across.
One detail decides whether the rest holds. Every row is written per cup of starting purée, which is 245 g. It is not written per cup of what you have left. A cup of blotted purée weighs less and carries more total solids. Total solids means everything that is not water: fiber, sugars, starch and minerals. Measure the drained purée by volume and your correction lands too small.
So: weigh, prep, weigh again, read the row. Three numbers, one line.
Falling between two rows? Take the lower one. Under-correcting a batter is easy to fix at the mixing bowl. Over-correcting is not.
Where the Five Levels Come From
Nobody invented these five levels. Each one matches a method bakers already use, converted to grams against the same 245 g cup. L1 is the can, opened and used, at roughly 90 % water by weight. L2 is one pass in a paper-towel-lined bowl. Sally’s Baking Addiction reports 340 g of purée coming down to 315 g that way. That is 7.4 % of the starting weight, or 18 g per cup.
L3 is three flat sheets, twenty minutes, left alone. It lands at 16 %, or 39 g per cup. L3 is the one modeled value on the scale, placed between the two measured points on either side of it. Every other level rests on a published figure. That matters for trust: you can check four of the five against their source.
L4 is pressing, then swapping in dry towels. Christina Marsigliese at Scientifically Sweet takes a third of a cup down to a quarter cup. That is 25 %, or 61 g.
L5 is the deep end at 50 %. You reach it by renewing towels until the purée stops giving, as in Sally’s crumb cake cookies. Or you reduce it in a pan. King Arthur Baking describes cutting purée by half in weight and volume, over ten to fifteen minutes.
Add Liquid Back, or Cut Flour?
Two corrections work, and they do not give you the same cake. Both corrections hold the ratio. They differ in what you end up with: the same batch with a steadier purée, or a smaller, denser one. Column A gives the batter back the water the purée lost, gram for gram. The yield stays put and the recipe behaves as written. Column B leaves the drier purée alone and removes flour instead.
The ratio between water and flour holds, but you get less batter and a tighter crumb. Pumpkin flavor comes through more concentrated, which some bakers want on purpose. Column A works at every level on this card. Column B works up to L3 and no further. The table below sets the two options side by side.
| A — add liquid back | B — cut flour instead | |
|---|---|---|
| Principle | Give the batter back the water the purée lost | Lower the flour so the water-to-flour ratio holds |
| Formula | add-back = water removed | flour cut = water removed ÷ 0.919 |
| Effect on yield | Unchanged | Smaller — less batter, fewer tins filled |
| Effect on crumb | The recipe behaves as written | Tighter crumb, more concentrated pumpkin flavor |
| Works up to | Every level | L3. Past that, the cut clears 25 % of the flour weight |
| Choose it when | You want your usual result from a steadier purée | You want a denser crumb on purpose |
How much liquid should I add back after draining pumpkin puree?
Exactly what you took out, by weight. Blot a cup down by 39 g, put 39 g of liquid back. Milk, buttermilk or water all do the job. Use whatever the recipe already contains. The batter returns to the hydration it was written for. That is column A, and it is the option that always works.
Do I need to change the flour if my pumpkin is drier?
Only if you want the drier result. Baker’s percentage — the habit of expressing every ingredient against flour at 100 % — makes the check easy. Our muffin base carries 240 g of flour and takes 91.9 % of its hydration from the purée. So the flour cut is water removed ÷ 0.919. That formula holds until it does not. At L4 it asks for 67 g out of 240 g. That is 28 % of the flour. What comes out of the tin is no longer the recipe you started with. Past a quarter of the flour weight, switch to column A.
Both columns were built on the muffin batter these numbers came from. The custard side has its own write-up: the pie column, explained.
What It Does to the Bake Time
Less water means less to drive off, so the bake finishes sooner. The effect is real, and it is small.
| Level | Muffins (min, from a 20 min base) | 9-inch pie (min, from a 55 min base) |
|---|---|---|
| L1 | 20.0 | 55 |
| L2 | 19.4 (−0.6) | 53 (−2) |
| L3 | 18.6 (−1.4) | 51 (−4) |
| L4 | 17.8 (−2.2) | 48 (−7) |
| L5 | 15.2 (−4.8) | 41 (−14) |
Does draining pumpkin change the baking time?
Yes, by minutes. The columns above come from a heat-load model built on the Siebel equation. That equation estimates a food’s specific heat from its water content. Specific heat is the energy it takes to warm something by one degree. The muffins are modeled at 75.7 g per cup, filled evenly. The pie is a 9-inch custard. Both columns are model outputs, not stopwatch readings, and both assume nothing else in the recipe moved. That assumption matters.
Change the tin, the oven or the fill depth and the model stops describing your bake. What survives is the direction and the size of the effect. Drier purée bakes faster, by a margin you measure in minutes. Not by a quarter of an hour, and not by half a bake. Treat the clock as a prompt, not a verdict.
Even at the extreme, the bake time moves by minutes. Draining your purée will not turn a 55-minute pie into a 20-minute one. If your pie is still loose after an hour, water is not the only thing going on. Check the oven and the probe.
Pull at temperature, using the doneness targets set out in the muffin and pie articles above. ThermoWorks publishes the reference ranges those targets rest on.
Two notes on scope. A homemade purée does not start at the same place on this scale. That is its own question: where a homemade purée sits on this scale. A dough that spreads on a flat sheet behaves unlike a batter in a tin. See cookies, where the same card applies differently.
Print It, Then Fill It In
The grid on this page is the whole thing. It is here in full, in text, with nothing held back for later. What the printed version adds is its back. Five blank rows, laid out like the table below, one line per bake. Cans vary by brand and by lot, and no published figure knows which one you bought.
After three bakes you stop reading an average and start reading your own store, your own brand, your own towels. That is the point of the blank rows. The card ships with our numbers. After three bakes it carries yours. A printed sheet also survives what a phone screen does not. Flour, wet hands, a timer already running. Tape it inside a cupboard door and it is there on the next bake.
| Brand / lot | Date | Weight before | Weight after | % removed | Recipe | How it turned out |
|---|---|---|---|---|---|---|
I did not set out to make a card. After four articles on the same purée, I noticed something. I was redoing the same arithmetic in my head before every batch. And getting it wrong whenever I was in a hurry. So I wrote the grid on a sheet of paper and taped it inside a cupboard door. That sheet became this card. What it still owes you is the tasting: these corrections come out of mass balance, not out of a panel. The only way to know for sure was to test it.
Print it if it helps. The file is a two-sided sheet, A4 and US Letter, readable in black and white. A short form on this page will email it over. The newsletter box beside it is separate, starts unticked, and the file arrives either way. Prefer to skip the form? Here is the direct link to the printable card.
Sources and Methodology
This card is a consolidation, and it says so. No new bench work sits behind it. Every figure is a published measurement, arithmetic run on published values, or a model output whose assumptions are named. The five water levels come from four published blotting and reduction protocols, converted to one unit. The two correction columns come from a mass balance on a 240 g flour base. The bake-time columns come from a heat-load model. Nothing here has been tasted against a control. That validation is scheduled, and this page will carry a v1.1 label when it lands. All twelve links below were tested on 23 August 2026.
- USDA FoodData Central — Pumpkin, canned, without salt (168450), SR Legacy. Water 90.0 g per 100 g; 1 cup = 245 g. Reference unit and total solids.
- Sally’s Baking Addiction — Soft Pumpkin Cookies, updated 16 April 2025. Light blot: 340 g to 315 g, 7.4 %. Level L2.
- Scientifically Sweet — Chewy Pumpkin Cookies, Christina Marsigliese, 11 October 2025. Three pressed paper towels, one third of a cup down to a quarter cup, 25 %. Level L4.
- Sally’s Baking Addiction — Pumpkin Crumb Cake Cookies, 3 October 2025. Renewed blot: 226 g to 113 g, 50 %. Level L5.
- King Arthur Baking — The simple trick for better pumpkin bakes, Kye Ameden, 4 October 2023. Reduction of ten to fifteen minutes, by half in weight and volume. Level L5, second route.
- King Arthur Baking — What is baker’s math?, Martin Philip, 28 April 2022. Flour at 100 %, everything else against it. Column B formula.
- King Arthur Baking — Ingredient Weight Chart, reference table. Flour 120 g per cup; purée 227 g per cup, against the USDA 245 g. First note under the card.
- Ankara University — FDE449, Physical Properties of Foods (PDF), open courseware. Siebel equation, c = 3.349 · mwater + 0.83736. Bake-time columns.
- ThermoWorks — Baked Goods Doneness Temps, updated 24 September 2025. Doneness ranges for quick breads and milk-rich custards. Pull-at-temperature note.
- Federal Trade Commission — CAN-SPAM Act: A Compliance Guide for Business, August 2023, revised January 2024. Header, identification, postal address and opt-out duties for the delivery email.
- Information Commissioner’s Office — Complying with the PECR electronic mail marketing rules, current guidance. Consent or soft opt-in for the newsletter box.
- Information Commissioner’s Office — What is valid consent? (UK GDPR), current guidance. Unticked box, unbundled consent, no conditioning of the file on it.
What to Remember
- Weigh the purée before you prep it and after. The difference is the only number the card needs.
- Put the water back if you want the recipe unchanged. Cut flour instead if you want a tighter crumb.
- Past a quarter of the flour weight, cutting flour stops being an option. Add liquid back.
- Bake times move by minutes, not by half hours. Pull at temperature, not at the clock.
Next step: print the card and tape it inside a cupboard door. Fill in the first line on your next bake. Then see how much water each brand starts with before you buy the next can.
Assisted by AI, reviewed by our human editorial team. View our Pages : Editorial Promise / Methodology / Disclaimer. This article is for informational purposes only and does not constitute medical or nutritional advice.
