Learning how to use a thermometer to check doughnut doneness takes about 30 seconds per doughnut and removes the guessing that ruins most home batches. You want the reading taken at the center of the dough itself, not the oil, and you want it to hit 190F or a little above. Everything else in this guide is about getting that one number without tearing the doughnut open or dropping a probe into 350F oil.
A little context on why the number matters: the outside of a doughnut browns long before the middle sets. By the time you see a deep golden ring, the center can still be wet, gummy batter. Oil color and the float test cannot see that. Only a probe can.
The short version: probe from the side, into the thickest part, read the center, and pair the number with a golden crust and a cooked crumb.
Table of Contents
What You Need for How to Use a Thermometer to Check Doughnut Doneness
You need two temperatures and two tools. The clip-on candy thermometer watches the oil so you never have to guess at the burner. The instant-read probe reads the dough. Trying to do both jobs with one tool is the most common reason the readings are wrong.
The short equipment list: a clip-on candy or deep-fry thermometer, an instant-read probe with a thin needle, tongs or a spider skimmer, a wire rack set over a sheet pan for draining, paper towels, and a heavy pot filled about two-thirds with a neutral oil.
Which thermometer type actually does which job
| Thermometer | Range | Mounts | Use it for |
|---|---|---|---|
| Candy thermometer | Up to about 400F | Clips to the pot rim | Holding oil between 350F and 365F |
| Deep-fry thermometer | Up to about 500F | Clips to the pot rim | The same job, with a wider dial and a face you can read from across the kitchen |
| Instant-read probe | Typically -50F to 550F | Held in hand | Reading the doughnut center and the oil surface |
| Meat thermometer with a thick probe | Varies | Held in hand | Oil only, if the probe is too wide to enter a doughnut |
A thin instant-read probe is the one that matters for doneness. Anything with a chunky stem tends to tear a fried doughnut open before it registers, and you lose the crumb you were trying to check.
Before the first fry, do an ice-water check. Put the probe in a glass of ice water with the tip submerged and held still. It should read 32F. If it reads 31F or lower, stir the water and repeat; if it still fails, the thermometer is off and every number it gives you later is off by the same amount.
Step-by-Step
Set Up the Thermometer and Heat the Oil
Clip the candy thermometer to the pot rim so the bulb sits in the middle of the oil, not against the side wall. A probe touching metal reads the burner flame and the pot, not the oil, and it will run high by 20 degrees or more.
Use Fahrenheit if you have a choice. The ranges people argue about, 350F versus 375F, only make sense on a Fahrenheit dial, and converting mid-fry is a good way to lose track.
Bring the oil to 350F for yeasted raised doughnuts and old-fashioned cake doughnuts, or 365F for cake-style and gluten-free batters, which are denser and need more heat to move. Then lower the heat so the temperature settles instead of climbing.
Keep the flame under the pot moderate. Big bubbles that roll across the surface mean you are on your way to 400F without noticing, and that is the range where doughnuts darken before they cook.
How to Use a Thermometer to Check the Center Without Tearing It
Insert the probe from the side, not the top. Push it horizontally into the widest part of the ring, angled slightly toward the center, until the tip sits in the middle of the thickest band of dough.
Stay away from three places. The center hole holds oil that will drip down the probe and give you a false low reading. The crust is already cooked, so a reading there is meaningless. And the top surface punctures easily, which is why a top-down poke tears a tender doughnut open.
Slide the tip in with one steady motion rather than a stab. A doughnut that splits on the way in tells you nothing, and you have wasted one from the batch.
Two details that trip people up. Doughnut holes and small pieces need the probe held from above and read instantly, because the window is seconds long. And a doughnut that is thick enough to hold a probe should be probed in a spot that will not show on the glazed face.
Read the Temperature at the Right Moment
Read the probe before you pull the doughnut out if you can manage it, or within five seconds of lifting it onto the rack. The number keeps climbing after the dough leaves the oil, so a late read is a late read.
Look for 190F as the floor for plain cake and old-fashioned doughnuts, and 195F for yeasted, brioche and filled doughnuts, which carry more fat and water through the middle. Above 200F the crumb turns dry and tight, and that is how you get the doughnut that tastes like it has been sitting out for an hour.
The curve inside a doughnut has three parts. The first thirty seconds or so, the center climbs fast from the temperature of the raw dough. Then it plateaus while the water in the crumb sets. Finally, carryover cooking adds another 5F to 10F on the rack, which is why a doughnut read at 188F in the pot finishes at 195F by the time you glaze it.
That plateau is exactly where the raw-center problem hides. Oil at 375F browns the outside in ninety seconds while the center is still climbing through its plateau. Oil at 330F holds the whole structure together longer but lets the dough drink oil, so you get a heavy, greasy ring instead.
For filled doughnuts, probe the plain ring on one side rather than the filled center. Jelly and custard hold heat differently from plain dough, and reading straight through a warm filling can push the number past the point where the surrounding crumb is actually set.
Confirm Doneness With Color and Texture
A reading is a checkpoint, not a verdict. Look for a uniform golden brown on both sides with a slightly darker ring at the edge, no pale dough showing near the hole, and a surface that springs back lightly when you press it with a finger.
The float test is a useful backup for thin doughnuts you cannot comfortably probe. A doughnut that floats and stays up is usually done, since the water inside has turned to steam. It is not proof, because thin doughnuts float early.
Signs the reading is unreliable: the number jumped more than 20F in two seconds, which means the probe touched the pan or the oil. Reset, wipe, and take it again. Also treat any doughnut that splits or streams oil onto the rack as unverified, and check a different one from the batch.
Internal Temperature and Fry Time by Doughnut Style
| Style | Oil temp | Internal target | Time per side |
|---|---|---|---|
| Yeasted raised | 350F | 190F to 195F | About 90 seconds |
| Old-fashioned / cake | 350F | 190F | About 2 minutes |
| Brioche | 345F | 195F | About 2 to 2.5 minutes |
| Gluten-free | 365F | 195F | About 2 minutes |
| Doughnut holes | 350F | 195F | About 45 seconds |
| Filled ring (probe the plain side) | 345F | 195F | About 2 minutes |
These are starting points, not rules. Your recipe, your oil, your altitude and your pan all move the numbers a little. Treat the table as the range you are correcting against after your first three doughnuts.
Common Mistakes
The raw center is the big one, and almost every case traces back to oil that is too hot for the batch size. A brioche doughnut that browned outside and stayed raw in the middle is the classic example: the oil was running far too hot, and the dough needed two full minutes to brown rather than ninety seconds.
| Symptom | What the reading shows | Cause | Fix |
|---|---|---|---|
| Raw center, dark outside | Center stuck under 185F while the crust browns fast | Oil above 370F, or too many doughnuts in at once | Drop to 345F, fry three at a time, wait for the oil to recover between batches |
| Greasy, heavy doughnut | Center fine at 190F, but fry time stretched past 3 minutes | Oil fell under 335F when the dough went in | Heat the oil a little hotter before each batch, and keep the batches small |
| Pale outside, done middle | 190F on a fast read, pale ring on the outside | Doughnut was pulled early or oil was under 340F | Let color drive the pull, then verify with the probe on the next one |
| Reading much higher than the recipe | Probe jumped quickly, plate reads the same | Probe touched the pot wall or the oil surface, not the center | Re-seat the probe in the middle of the oil and the thickest dough |
| Dry, tight crumb | 205F or above | Doughnut left in the oil past the target | Pull at 190F and lean on carryover cooking to finish it |
Batch size is the lever people miss. Oil temperature drops the moment dough goes in, and three doughnuts at a time is about the practical limit for a home pot. Drop six in and the temperature falls far enough that you absorb oil and then crank the burner to catch up, which scorches the outside before the centers catch up at all.
Never leave the probe sitting in the oil while you fry. Clip-on for the pot, handheld for the dough, and swap only when you are actually reading.
When the oil overshoots, do not fight it. Turn off the heat, wait two minutes, then add a small ladle of cool oil to bring it back down. Adding cold oil to a very hot pot spatters more than the temperature drop is worth.
For safety, dry the doughnuts thoroughly before they go near the oil, lower them away from you with a spider or tongs, and never fill the pot more than halfway. A pot that is two-thirds full has room to drop dough without a wave over the rim.
And keep a lid within reach, a fire extinguisher in the kitchen, and never leave a fryer unattended. A pot of 365F oil on a gas burner will go from fine to smoke in the time it takes to answer the door.
Frequently Asked Questions
Can I use a meat thermometer to check doughnuts?
Yes, as long as the probe is thin enough to enter the dough without splitting it. A slim instant-read probe is ideal: slide it in from the side of the ring into the thickest part of the dough and read the center, staying clear of the hole and the crust. Thick meat thermometers work for checking oil temperature but are usually too wide to probe a tender doughnut without tearing it open.
Should I check the oil temperature or the doughnut temperature?
Both, because they answer different questions. The clip-on thermometer tells you the oil is holding 350F to 365F, which sets the crust and controls how much oil the dough drinks. The instant-read probe tells you the doughnut center has reached 190F or more, which is the only way to know the inside is set. A doughnut can brown beautifully in oil that started too cool and still be perfect, or brown in 380F oil and stay raw inside.
Where exactly do I put the probe in a doughnut?
Insert it from the side, horizontally, into the thickest band of dough, angling slightly toward the center until the tip sits in the middle. Avoid the center hole, where trapped oil will drip down and drag the reading down, and avoid the crust, which is already cooked. Push in with one steady motion rather than a stab, or the doughnut will split and give you nothing to read.
Can I check a filled doughnut without it bursting?
Yes, and you should not probe the filled side. Push the probe through the plain ring on the opposite side, into the thickest part of that band. Custard and jelly hold heat differently from plain dough, so a reading taken straight through a warm filling can climb well past the point where the surrounding crumb is actually set. Fry the filled doughnuts slightly cooler, around 345F, so the shell does not outrun the filling.
Does a donut style change the internal doneness target?
A little. Plain cake and old-fashioned doughnuts are done at about 190F, while yeasted, brioche, and filled doughnuts usually need 195F because they carry more fat and water through the middle. Gluten-free batters are denser and benefit from the same 195F target with hotter oil. Past 200F the crumb turns dry and tight, so read early and let carryover cooking finish the last 5F to 10F on the rack.
Conclusion
Start with the tool. Put the probe in ice water, confirm it reads 32F, and clip the candy thermometer into the middle of the oil rather than against the pot.
Then take one reading on the first doughnut of the batch: in from the side, into the thickest part, away from the hole and the crust. 190F for cake and old-fashioned, 195F for yeasted, brioche and filled.
Compare that number against the color, the spring, and the time in the oil. When all three agree, you have a doughnut you did not have to cut open to find out about.
