What temperature should your fridge actually be set to?

The short version: the food safety target is 4°C (40°F) or colder for the fridge, and -18°C (0°F) or colder for the freezer. Almost no dial inside an actual fridge shows either number. It shows 1 to 5, or a bar between MIN and MAX, and that gap between the number you can turn and the number that actually matters is the reason most people have never checked what their fridge is really running at.

Why the dial does not show a temperature

Most fridge thermostats are still built around a simple mechanical part: a sensor that reacts to temperature and a dial that adjusts how sensitive it is, not a dial that sets a target in degrees. Turning it to 5 does not mean "5 degrees", or "colder than 3" in any way you can count on across brands — it means "the compressor runs longer before switching off", and how that translates into an actual shelf temperature depends on the model, how full the fridge is, and the room it sits in. Two fridges set to the same number 3 can easily be running two or three degrees apart. The number on the dial was never designed to be compared between appliances, or even trusted on its own — it is a relative adjustment, not a reading.

The number that actually matters, and why it is that one

Food safety guidance converges on 4°C (40°F) as the ceiling for a fridge because of how bacteria behave, not because it is a round number. Growth is slow well below that point and accelerates sharply once temperature climbs into what is generally called the danger zone, roughly 4°C to 60°C (40°F to 140°F) — inside that range, bacteria that cause illness can double in number in well under an hour under the wrong conditions. 4°C is not the point where growth stops entirely; it is the point below which it is slow enough that the shelf-life estimates on packaging, and the judgement calls covered elsewhere on this site, actually hold. A fridge running at 7°C instead of 4°C is not a little bit warmer. It is a fridge where everything's clock is quietly running faster than the label assumed.

The freezer's -18°C target works on a different mechanism. It does not need to kill anything — freezing was never a sterilising process — it needs to stop water from being liquid enough for bacteria to use, which halts growth almost entirely rather than slowing it down. That is also why freezer temperature drifting a few degrees warmer is a slower-motion problem than the same drift in a fridge: growth does not really resume until things approach 0°C, but the ice-crystal damage behind freezer burn does get worse the closer the freezer runs to that point.

One dial, several actual temperatures

Even a fridge that is genuinely holding 4°C by the sensor is not 4°C everywhere inside it — which shelf gets which temperature, and why the door runs warmest, is its own question with its own answer elsewhere on this site. What is worth adding here is that the dial controls roughly the middle of that range, not any specific shelf, so "the fridge is set to 3" tells you almost nothing about whether a specific carton on a specific shelf is actually at or under 4°C.

How to actually check, instead of guessing

The only reliable fix is a separate thermometer, and it is a small, one-time thing to sort out. An inexpensive fridge thermometer, left on a middle shelf toward the back — not the door, not pressed against the back wall where it can misread — for several hours, tells you the real number for the first time. If it reads above 4°C, move the dial up one notch and check again a day later rather than adjusting repeatedly in one sitting; the fridge takes time to settle after each change, and judging it too soon just leads to overcorrecting. The same thermometer works for the freezer, aiming for -18°C or colder.

Worth rechecking seasonally rather than assuming a setting that worked in winter still holds. A warmer kitchen in summer, a fridge pushed harder by more open-door moments during a holiday, or simply a fuller fridge holding more cold air in place, all change what a given dial position actually delivers.

The sign it has drifted, even when nothing looks wrong

A fridge running a couple of degrees warm rarely announces itself. Nothing frosts over, nothing smells off on day one, and the door still swings shut the same way it always did. What actually shows up is that several unrelated things start turning a little earlier than expected in the same stretch — milk souring sooner than usual, leftovers on the edge a day early, greens going limp faster than the last bag did. One early item is probably that item. Several unrelated ones turning early in the same week is a fridge worth checking with a thermometer before anything else.

What actually helps day to day

• Buy an actual thermometer once. It costs very little and answers a question the dial was never built to answer.
• Target 4°C (40°F) or colder for the fridge, -18°C (0°F) or colder for the freezer. These are the numbers that matter, not the notch on the dial.
• Adjust one notch at a time, then wait a day. A fridge needs time to settle before the next reading means anything.
• Recheck after a season changes, or after a big shop fills the fridge. The same dial setting does not deliver the same temperature all year.
• If several things spoil early in the same week, suspect the fridge before you suspect yourself. That pattern points at temperature, not memory.

Getting the temperature right buys every single item in the fridge a few extra honest days — but it does not do anything about noticing which item needs eating first, which is a separate problem with a separate answer. That is the part Munch is built around: tracking what is actually in the fridge and what is closest to its own limit, shared with everyone in the household, so the few extra days a correct fridge temperature buys you do not just get spent on something nobody remembered was there.

None of this needs a smart fridge or a lucky guess. It needs one cheap thermometer, checked once properly, and revisited whenever the season or the shopping changes enough to matter.