In this article
A guest rarely complains about CO2 by name. They say "it felt stuffy in there", or they say nothing and just skip the second bottle, skip dessert, skip the coffee after — and leave fifteen minutes earlier than planned. The room was quiet, the thermostat was fine, and it still felt heavy. That isn't ambiance. That's air that has been circulating for a whole service without being replaced.
CO2 itself isn't toxic at the concentrations a dining room ever reaches — the problem isn't the gas, it's what it's a proxy for. Every breathing guest exhales it, and the rate at which the concentration in the room climbs is exactly the rate at which the room is NOT being refreshed. A high CO2 reading is therefore a direct measurement of how much fresh air a guest is actually getting between breaths — and research finds that number tracks directly with how "fresh" or "stuffy" a room feels, independent of temperature or noise.
Outdoor air sits around 420 ppm CO2. A well-ventilated room stays close to that. From roughly 1,000 ppm, a meaningful share of people in a space start reporting the air feels "stale" or "heavy" — a century-and-a-half-old building-science rule of thumb that still holds up today. And at 2,500 ppm, a widely cited study measured a genuine, measurable drop in decision-making ability in people who were simply sitting there working.
Those are the two thresholds in the title, and they weren't picked at random — they're the two points where research finds a kink in how people actually behave. Underneath it is also the reason two rooms playing the exact same music at the exact same volume can feel completely different: not the seat count, but how much fresh air arrives per seat.
Below is a calculator: enter your room, your covers and your ventilation, and see which band you land in by the end of a service — and what one realistic upgrade would move it to. Everything runs in your own browser; nothing is sent or stored.
What the research actually says, in numbers
The rule of thumb that air above 1,000 ppm starts to feel stuffy comes from the German chemist Max von Pettenkofer, who proposed it back in 1858 as a practical marker of "adequate ventilation" in an occupied room. The US ventilation standard ASHRAE uses a similar principle today: keep indoor CO2 within roughly 700 ppm of outdoor air — at an outdoor baseline of 420 ppm, that works out to an indoor ceiling of about 1,100 ppm. Above that line, ASHRAE's own guidance notes that more than 20% of occupants report discomfort with the air quality, and that discomfort climbs further as the concentration rises.
The second number comes from one specific, widely cited experiment. Researchers at the US Lawrence Berkeley National Laboratory (Satish et al., 2012) put twenty-two participants through the same decision-making tests in rooms held at 600, 1,000 and 2,500 ppm CO2. At 1,000 ppm, average scores were 12% lower than at 600 ppm, with a noticeable drop on six of nine measured skills. At 2,500 ppm, that fell to 51% lower, with scores the researchers themselves called "dysfunctional" on initiative-taking and strategic thinking — two things a host or a server needs all night, not just a guest.
One thing worth knowing: ASHRAE's own standard doesn't set a maximum ppm as a building requirement — the standard specifies a ventilation FLOW RATE (litres of outdoor air per person per second), not a concentration ceiling. That's exactly why code-compliant ventilation and a room that feels fresh are two different promises: the law says how much air comes in, not how the air feels by the end of a busy service. The two only line up if the flow rate actually scales with how full the room is.
Five things that decide the air in your room
In order of how quickly you can act on them: the first two cost an evening, the last two cost a conversation with your installer.
1. Measure it, instead of guessing at it
Nobody can feel the difference between 900 ppm and 1,400 ppm — the sense that "it's heavy in here" usually only kicks in well past the point where it had already been a problem for a while. An indoor CO2 monitor now costs about the same as a couple of bottles of wine, and it replaces a hunch with a number.
Put it at table height, in the middle of the room, away from a door or a vent — those two spots measure the draught, not the air your guests are actually breathing. Read it at three moments: opening, the busiest point of service, and just before close. That third figure is usually the most telling one, because it's the air the last guests of the night are sitting in.
One evening of measuring is enough to know which of the three bands below your room lands in on a normal busy night. After that you'll know whether the rest of this guide is a matter of habits, or a matter of the installation itself.
Outdoor air as the zero point, and the two thresholds research actually finds a kink at — not the hundreds of ppm in between, but the two points where people's behaviour changes.
The two bold thresholds come from published research (see above); the values in between are illustrative. Your own CO2 monitor gives the only reading that matters for your room.
2. Ventilate BEFORE service, not during it
CO2 doesn't spike the moment the first guest walks in — it builds up, and how fast depends on how much air your room replaces per hour. In a moderately ventilated room that build-up window is often twenty to thirty minutes. Which means the window to do something about the main-course air has already passed a full half hour before the doors open.
The practical rule: run mechanical ventilation (or windows and doors, where that's an option) at full capacity a half-hour before opening, rather than only once the room is already full and the reading is already climbing. By the time a system "catches up" at full speed, the first guests have already spent an hour sitting in last hour's air.
The same applies between services. A room that doesn't get a proper flush between lunch and dinner starts the evening service carrying the CO2 buildup from the afternoon. A quarter-hour of full ventilation between services is the cheapest reset there is.
3. Your kitchen extraction pulls its air from somewhere — check where
A hood over the stove pulls air away, and that air has to come from somewhere: that's make-up air, and it's exactly the subject of the kitchen extraction and air balance article. If that make-up air isn't deliberately drawn in from outside, the system pulls it from whichever nearby space has the lowest pressure — and that's often the dining room.
The result is a double hit at the busiest moment of the night: exactly when the kitchen is running hardest and pulling the most air, that's exactly when the most air gets pulled out of the dining room — the room that, at that same moment, also holds the most guests exhaling the most CO2. A room that feels fine during the day can specifically get worse during the evening peak because of this.
The question for your installer is one sentence: "does the make-up air for the kitchen extraction come directly from outside, or does the system pull it from the dining room?" On the second answer, you're solving the kitchen at the expense of the room your guests are sitting in.
4. Match ventilation to your peak covers, not your average
A system sized for an averagely full room is, by definition, undersized on every night that's busier than average — and those are exactly the nights it matters most. Ventilation rate should scale with the number of people in the room, not with the floor area alone.
Below is how four common situations compare, and roughly how much fresh air per person each one typically delivers.
The US ventilation standard ASHRAE 62.1 uses a benchmark for a restaurant dining room of about 7.5 cfm per person (roughly 3.5 litres per second) plus a small allowance per square foot for cooking odours. That's a legal minimum, not a comfort promise — run that minimum for your own room and a full room on a long service often still finishes well above the 1,000 ppm line, even while technically meeting the standard.
How much fresh air per person per second each situation typically delivers. Higher is better — and it drives directly how high CO2 climbs by the end of a service.
ASHRAE 62.1's own benchmark for a restaurant dining room works out to roughly 3.5 to 5 litres per person per second, depending on how tightly the room is seated — that's the US legal minimum, not a guarantee of a fresh room. "High-capacity" above is the level at which a full room stays under the 1,000 ppm line even by the end of a long service.
5. Read the signals your guests are already giving you
Guests rarely say "CO2", but they do say something. "It was warm and stuffy in there" in a review, while the thermostat was fine. A table that skips dessert and asks for the bill early. Fewer second bottles of wine and less coffee after the main course, with nothing changed on the menu or the price — all signals that could just as easily point at heavy air as at slow service.
The distinction is testable: does the first twenty minutes of a service feel different from the last hour? Do you notice a difference between a quiet Tuesday and a full Saturday, at the same temperature and the same music? If the answer to both is yes, that points more strongly at air than at atmosphere — acoustics and temperature don't naturally change with how busy it gets, ventilation load does.
Tie it to your own reviews: a search of the past year's Google and Tripadvisor reviews for "stuffy", "stale", "warm and airless" or "no fresh air" often gives a more honest answer than a hunch. If those words cluster on your busy nights, that's reason enough to take the calculator below seriously.
Run the numbers on your own room
Enter your covers, your room and your ventilation type. The calculator builds up CO2 over the length of your service using the same formula used in ventilation engineering, and shows which band you land in — plus what one step up in ventilation would change.
This is deliberately a rough estimator, not a substitute for an actual air-quality measurement: it assumes an average CO2 output per seated, talking guest and typical ventilation rates per category. Your own CO2 monitor (step 1 above) remains the real answer for your room.
Freshness estimator
Room, covers, service length and ventilation type → an estimated CO2 reading by the end of service.
—
—
The model assumes an average CO2 output of 0.0075 litres per second per seated, talking guest (a common building-services estimate) and the ventilation rates above. Outdoor air is set at 420 ppm. Everything runs in your own browser; nothing is sent or stored.
Two things worth remembering when reading the result. The model computes towards a steady-state value that depends on your ventilation rate PER PERSON — not on the number of seats by itself. A big room with weak per-person ventilation can therefore end up worse than a small room with a properly sized system, even though "big and airy" sounds better on paper.
That's exactly why the legal minimum and a fresh room are two different promises: a system that exactly meets the standard meets a flow-rate requirement, not a ppm promise — and the two only line up if the flow rate genuinely scales with how full the room gets on your busiest night.
What to do with this this week, this month and this quarter
Tackling five things at once works for nobody. This order works, because each step tells you whether the next one is worth doing.
This week — measure once
- Buy an indoor CO2 monitor and place it at table height, away from doors and vents.
- Read it at opening, at the busiest point, and just before close, on your two busiest nights.
- Enter your own numbers into the calculator above and compare the estimate against what your monitor says.
- Search your own Google and Tripadvisor reviews from the past year for "stuffy", "stale" and "warm and airless".
This month — change habits, not the installation
- Run ventilation at full capacity a half hour before opening rather than only once the doors open.
- Fully flush the room for fifteen minutes between lunch and dinner service, even in cold weather.
- Ask your installer where the make-up air for the kitchen extraction actually comes from.
- Repeat the measurement on your busiest night after these two habit changes and compare the difference.
This quarter — if the habits aren't enough
- Get a quote for higher ventilation capacity if your busiest-night reading still sits above 1,500 ppm after the habit changes.
- Work out the legal minimum for your own room (7.5 cfm per person plus a per-square-foot allowance) and compare it against what your installation actually delivers.
- Add the measurement to your regular routine alongside your other room checks, like temperature and acoustics — the three together decide how a room physically feels.
- Share the number with your team: "stuffy" becomes a measurable signal instead of a vague feeling nobody names first.
The air is the sense nobody tests
Temperature is felt within a minute. Noise is heard from the first sentence. Air builds up over an hour, and that's exactly why it's never the first suspect when a room "just doesn't feel right" — even though it's often the actual reason.
The good news is the fix is rarely an expensive installation. Starting ventilation half an hour earlier, flushing the room for fifteen minutes between services, and knowing where the kitchen's make-up air actually comes from solve most of the problem at most venues — for the price of a habit, not a renovation.
Do the same next with the other two senses a guest rarely names but always feels: your room's temperature and its acoustics. Three physical senses, three numbers instead of a feeling — which is exactly what a guest means when they say a room "just works".