Scope. This page provides a controlled reader-run protocol and does not claim site-run measurements. For cooking, cooling, storage and reheating, follow the current official guidance linked on the page.
"Rest it for ten minutes" is one of the few pieces of cooking advice that has achieved total consensus. It appears in almost every recipe involving a piece of meat, it is repeated with confidence, and it is stated without any of the three variables that determine whether it is correct: how heavy the piece is, how hot it was when it came off the heat, and how cold the room is.
A quoted rest time with no mass attached cannot be right for both a single portion and a joint. The same instruction applied to both will overshoot one and undershoot the other, and the reader who follows it and gets a poor result will conclude they did it wrong.
This post replaces the quoted minutes with a way of measuring your own curves, and is explicit about which of the claims made for resting a home measurement can support and which it cannot.
What A Rest Is Supposed To Achieve
The usual advice bundles together several claims that deserve to be separated, because they are not equally well supported and they do not all point to the same rest time.
- Carryover. The interior of a piece of meat continues to rise after it leaves the heat, as energy stored in the hot outer layers moves inward. This is uncontroversial and is a straightforward consequence of a temperature gradient existing inside the food. It is measurable, and how to measure carryover cooking shows the method.
- Juice redistribution. The most commonly repeated claim, usually phrased as juices "redistributing" or "relaxing back into" the meat, with the observable consequence that a rested piece loses less liquid to the board when cut.
- Temperature evening. The gradient between a hot exterior and a cooler centre flattens during a rest, which changes the eating experience across a slice.
- Ease of carving. A commonly stated practical benefit, rarely with anything attached to it.
Those are four different claims. They are usually delivered as one sentence.
What to measure
The test design follows the variables the advice omits.
| Variable | How it is handled |
|---|---|
| Mass | Several masses, spanning a portion-sized piece to a joint, each logged separately |
| Starting internal temperature | Fixed target at removal, verified with a calibrated probe |
| Ambient temperature | Measured and logged for every run; the kitchen is not assumed to be a constant |
| Covered versus uncovered | Run as a paired comparison on matched pieces, foil-tented and bare |
| Surface | Same board, same orientation, same air movement |
| Time | Continuous logging from removal until the curve has clearly turned and continued down |
| Liquid released on cutting | Weighed, at several rest durations, against an unrested control |
For every run, record the probe used and its calibration offset, and the replicate count, including where the replicate count is one.
This page gives no curves and no liquid weights; record your own, with the ambient conditions for each run.
The Temperature Curves
Do not assume a particular peak or duration. Log core temperature at fixed intervals and report the observed maximum, time to maximum and subsequent decline alongside the cut's mass and geometry.
What to record, per mass:
- Time from removal to peak internal temperature.
- Magnitude of the rise from removal to peak.
- The slope of the decline afterwards, which is the part almost nobody measures and the part that determines whether dinner is warm.
- The same three figures for the covered and uncovered condition side by side.
- Both temperature scales throughout, Celsius canonical.
Keep the full curve for every run, not only the figures above.
Mass is the dominant variable and this is why no single time can be correct. A large joint has a small surface area relative to its volume, so it holds heat for a long time, peaks later, and declines slowly. A small piece has a large relative surface area, peaks quickly and then loses heat quickly. Any rest instruction that does not name a mass is, at best, correct for one size of thing.
Covered Versus Uncovered
This gets its own measurement because it changes two things that pull in opposite directions.
Covering slows the decline. A foil tent reduces radiative and convective loss from the surface, so the piece stays hot longer. If your problem is serving cold food, this helps.
Covering also softens a crust. Trapping vapour against a dry, browned surface rehydrates it. Anyone who has tented a roast and then found the crackling or the crust disappointing has observed this directly.
So the honest answer is conditional, and the condition is what you care about more on the plate. Record both curves, photograph the crust, and choose by what you care about. We give no covered-versus-uncovered figure for any mass; your paired runs supply it.
A loose tent, not a wrap, is the version that most people mean when they say "cover it". Wrapping tightly is a different thing entirely and behaves accordingly.
What Genuinely Changes During The Rest
Use the resulting trace only for the tested cut and conditions. The durable conclusions are:
Carryover is real and measurable. The interior continues to climb after removal. This is why the pull temperature is the number that matters and the target temperature is what you get. If you are cooking to a target and pulling at the target, you have overshot, and the size of the overshoot depends on mass and on how aggressively you cooked.
The temperature gradient flattens. A piece removed from a hot oven has a substantial difference between its outer and inner temperature. That difference reduces during a rest. Whether the reader experiences this as an improvement is a matter of preference, but it is a physical change and you can log it.
Liquid released on cutting changes with rest time. This is measurable directly by weighing what ends up on the board, so measure it at several rest durations against an unrested control.
What A Home Measurement Cannot Verify
This section exists because it is more useful than the previous one.
This experiment does not claim a mechanism. It measures mass before cooking, liquid released during the rest, liquid released on cutting, and temperature at each stage. Those observations can establish whether resting changed liquid loss in this setup without requiring a story about “redistribution.”
"Rests for as long as it cooked" and similar rules of thumb. These circulate widely with no supporting measurement we could locate. Check whether your curves are consistent with them and stop there.
Ease of carving. Frequently claimed, rarely quantified, and difficult to measure honestly without a texture rig. Describe what you observe and label it as an observation rather than a result.
If one of the claims above turns out to be unsupported by your data, record that as well. Keeping the ones that failed is the point.
The Trade-Off Nobody States
Here is the thing missing from almost every rest instruction ever written: a rested piece of meat is a cooling piece of meat.
The advice tends to be delivered as though the rest has only benefits and the only cost is patience. In practice a small piece resting on a cold board in a cool kitchen is losing heat quickly, and the reader who follows a generic instruction for a portion-sized piece is often serving something noticeably below the temperature they wanted. That is a real, common failure, and it is the direct consequence of applying a joint's instruction to a steak.
The mitigations are all straightforward once the trade-off is named:
- Warm the plates and the board. The single largest and cheapest intervention. A cold plate takes a surprising amount of heat from the food the moment it lands.
- Scale the rest to the mass. Small pieces need a short rest; large ones can take a long one and lose little.
- Tent when heat retention matters more than crust, and leave it bare when the reverse is true.
- Rest somewhere that is not a draught. Air movement across the surface accelerates the decline more than most people expect.
One deliberate omission: this post gives no guidance whatsoever on how long cooked food can be held out of refrigeration. That is a food-safety question, it is answered by named authorities rather than by us, and the named authorities are linked in how long does it actually keep. We are not going to improvise a number here, and no cook should accept one from a resting article.
In Plain English
What this means. Rest advice is quoted in minutes without stating mass, starting temperature or room conditions, which are the three things that determine the answer. Meat keeps climbing after it comes off the heat, then cools, and how fast both happen depends mostly on how big the piece is.
Who this helps. Anyone who followed the standard advice and served something either overcooked or lukewarm, and anyone cooking to a target temperature who has not adjusted their pull temperature for carryover.
When to use it. At the point you take the food off the heat, which is earlier than most recipes imply.
What to do next. Work out your own carryover for the size of thing you cook most often, and warm your plates.
Set Your Own Pull Temperature
The pull temperature that follows from all of this depends on your mass and your method, so take the pull point from a rise you have measured on your own equipment rather than from a memorised rule of thumb. If you are cooking at very low oven temperatures, note that carryover behaves quite differently and is much smaller, which is covered in low-temperature roasting.
Related technique: What Eight Hours Actually Does To A Piece Of Beef Shin.