I set three ovens to Gas Mark 4 and put the same thermometer in each one. The averages came out at 182, 168 and 182 °C. Here is what the dials were hiding.
Gas Mark 4 means 180 °C. That is the standard, and all three ovens in this test were set to exactly that mark. The same thermometer, on the same middle shelf, then told three different stories: my own oven averaged 182 °C, my mother's older cooker averaged 168 °C, and a bakery oven I was allowed to test on a slow morning averaged 182 °C. Two dials were honest. One was politely lying by twelve degrees, and it belonged to the best baker in the family.
Here is the whole test on one table, before the stories behind each number.
Oven
Setting
Average over 40 minutes
Lowest reading
Highest reading
My oven
Gas Mark 4
182 °C
171 °C
193 °C
My mother's cooker
Gas Mark 4
168 °C
157 °C
181 °C
The bakery oven
Gas Mark 4
182 °C
176 °C
188 °C
Read the averages first. An oven lives at its average, because the burner cycles on and off and the air temperature rides up and down around the middle value. The swing matters too, and I will come back to it. The headline, though, is the twelve-degree gap in the middle row.
How I Ran The Test
One thermometer, three kitchens, one protocol. Each oven was lit from cold, set to Gas Mark 4, and left alone for twenty minutes. Then the thermometer went onto the middle shelf, centred, and I wrote down a reading every five minutes for forty minutes without opening the door in between. No food, no trays, no helpful adjustments. At the end I averaged the nine readings and noted the lowest and highest.
The protocol matters because ovens behave differently in the first minutes, and a single peek at minute ten convicts or acquits a dial unfairly. The standard the dials claim to follow sits in the full Gas Mark conversion table, where Mark 4 is 180 °C conventional and 160 °C in a fan oven. Every oven in this test was a conventional gas oven, so 180 °C was the promise all three had made.
One more ground rule: I tested the ovens as their owners actually keep them. My mother's cooker has cooked daily for years and wears its age in its seals. Cleaning, servicing, or favouring one oven would have made a tidier experiment and a useless one. Readers do not bake in laboratory ovens. They bake in family ovens, with family wear on them.
Oven One: Mine Overshoots, Then Behaves
My oven started like an eager apprentice. The first reading, five minutes after the thermometer went in, was 193 °C, thirteen degrees over target. The burner had run hard through the warm-up and the air had carried the momentum with it. Then the readings settled into a rhythm: 186, 179, 184, 180, 188, 182, 171, 186. Hot breath, pause, hot breath. The average of 182 °C sits close enough to the promise that I have stopped worrying about this dial.
The swing, from 171 to 193, is wider than it looks on paper, and it explains a small mystery in my own baking. Anything that goes in during a hot crest browns a touch faster on top. Anything that goes in during a dip needs a minute or two longer. Across a forty-minute cake the two cancel out. Across a twelve-minute tray of biscuits they do not, which is why my second tray so often looks calmer than my first.
A dial is a request, not a measurement. The thermometer is the measurement.
Oven Two: My Mother's Cooker, Twelve Degrees Cool
My mother's cooker climbed slowly and never arrived. Readings: 163, 170, 166, 172, 168, 175, 169, 181, 157. The one reading of 181 °C looked like a correction until the next reading fell back to 157. The average, 168 °C, is twelve degrees below what the dial promises, and the swing of 24 degrees is the widest of the three ovens by a distance.
Then the family history snapped into focus. My mother's cakes have always needed ten minutes more than the book says. Her roasts rest while the potatoes catch up. She has adjusted for this oven for years without naming it, the way you lean against a familiar slope. When I showed her the average she laughed and said the oven had been like this since a repair years ago, and that she simply cooked by looking, not by the book.
Her looking, it turns out, is a calibrated instrument. Her dial is not. Every "ten extra minutes" in her recipes was that twelve-degree lie, being quietly paid for, one bake at a time.
Before I packed the thermometer away, I ran one confirmation bake. The same sponge mixture, split between two tins, one in her oven at her usual Gas Mark 4 and one at Gas Mark 4 plus a nudge toward the next half mark. The nudged tin rose evenly and left the oven at the book's time. The plain tin needed its customary ten extra minutes and came out paler at the edges. The twelve degrees were not a theory. They were lunch, delayed, for years.
Oven Three: The Bakery Oven Earns Its Keep
The third oven belongs to a small bakery that lets me buy day-old rusks and, on one slow morning, borrow twenty square feet of workspace. Their oven is older than mine and works far harder: it is lit before dawn and holds heat through a full production day. Its readings were the dullest of the test: 179, 184, 181, 186, 180, 182, 188, 176, 184. Average 182 °C, swing of only 12 degrees.
Temperature swing at a Gas Mark 4 setting, forty minutes per oven (°C)
Data: my thermometer log from the three test sessions described on this page. Each candle spans the lowest to highest reading; the body spans the first to the last.
The baker was not surprised by any of this. His thermometer lives in the oven permanently, hooked on the middle rack, and he glances at it the way a driver glances at a speedometer. He also made the point that a heavy, often-used oven holds a steadier heat than a light oven used twice a week, because the metal itself stores warmth between burner cycles. My mother's cooker, light and elderly, swings because there is less oven to smooth the ride.
What A Swing Or A Shortfall Does To A Bake
A shortfall is the worse fault. Twelve degrees low steals rise from cakes, stretches every timing in the book, and leaves chicken skin pale at the moment the meat is done. A swing is sneakier.
It averages out across long bakes and ambushes short ones, which is why a cook can produce a perfect Sunday roast in an oven that burns every tray of biscuits. My own worst version of this was a pizza base scorched at a corrected-for-fan setting, a story I told in full when I Set Gas Mark 6 In A Fan Oven and read the aftermath off a thermometer.
The practical reading of this whole test fits on an index card. Trust the average, not the dial and not a single reading. Forgive a swing of ten degrees either way. Investigate any average that lands more than ten degrees from the mark, because at that distance the recipes stop being wrong and the oven starts being wrong.
The Use Case: Cakes That Fail In A Pattern
This test is for the reader whose bakes fail in a pattern rather than at random. Random failure points at ingredients, weather, or mood. Pattern failure, the same paleness, the same extra ten minutes, the same sunken middle in the same oven, points at the dial. One thermometer and one free hour separates an oven problem from a cook problem, and the two problems have very different prices. My mother spent years assuming she had slowed down as a baker. She had not. Her cooker had.
What I Changed After The Test
Three habits survived the test. The thermometer now lives in my oven full time, hooked where I can read it through the door glass. My mother's recipes, the ones written against her cooker, carry a pencil note in my copies: her Gas Mark 4 is my Gas Mark 4 with a breath more, and her timings run ten minutes long. And I stopped trusting the warm-up light entirely, a distrust that later pushed me to time the warm-up itself, which is its own story in My Stopwatch Test Of Preheating To Gas Mark 5.
The dials were never really doing what they claimed. They were doing what their age, their seals, and their metal allowed. Once I knew each oven's actual behaviour, the same three dials became honest tools again, because I had finally read them properly.
Questions People Ask About This Page
Is A Temperature Swing Around The Setting Normal For A Gas Oven?
Yes. A gas thermostat works by cycling the burner on and off, so the air inside swings around the average instead of sitting still. In my test the widest swing was 24 degrees, in my mother's older cooker. Judge an oven by its average across half an hour, not by a single reading.
My Oven Runs Cool, Like The Second Oven In This Test. What Should I Do?
Treat the dial as a starting guess and cook to the thermometer for anything that matters. Half a mark up brought my mother's cooker close to a true 180 °C average, and her cakes stopped needing the extra ten minutes nobody could explain.
Where Should An Oven Thermometer Sit During A Test Like This?
Middle shelf, as close to the centre as the racks allow, clear of the walls and not sitting over the burner slots. A thermometer touching metal or hanging in a corner reads the corner, not the oven your food cooks in.
Should I Repair An Oven That Runs Twelve Degrees Cool?
Start with the thermometer. It costs little and tells you within one bake whether the dial drifts by a few degrees, which most ovens do, or by twelve or more, which points to a tired thermostat worth a repair visit.
Talha Kayani is the founder of Trivexa Digital, Multan, and the owner and developer of Ti-84calculator.com. A digital marketing specialist in SEO, social media and Meta Ads, he builds and personally tests every calculator and converter on this site before it goes live, so the figures you read here are the ones he trusts in his own daily work.
Written by 
Reviewed by 
