The Recirculation Button: Why Long Drives Make You Sleepy, and Why Your Restomod Has It Worse

The recirculation button closes a door. That is all it does. It shuts the flap that lets outside air into your HVAC box, and from that moment the system stops pulling fresh air and starts running the same cabin air in a loop. Everyone in the car keeps exhaling into it. On a long drive with a full car, that air gets measurably worse.

Whether that is why you feel like garbage two hours into a road trip is a more interesting question than the internet lets on, and the answer for a classic is different from the answer for a 2026 crossover. If you have built a car with new weatherstrip, one-piece door glass and bonded flush-mount windshields, you have made your cabin tighter than it left the factory. That is a good thing. It also means the fresh air path is now something you have to think about on purpose.

What does the recirculation button actually do?

It moves a single flap, usually called the recirc door or fresh air door, inside the heater and evaporator box. Open, the blower draws outside air in through the cowl. Closed, the blower draws air from inside the cabin instead. Nothing else changes. The compressor, the blower speed and the temperature blend all work exactly the same either way.

The reason it exists is efficiency and isolation. Cabin air is already cooled, so recycling it means the evaporator has less work to do and the car gets cold faster. It also keeps out whatever is happening outside, which in Phoenix in July is a real feature.

The icon is usually a car outline with an arrow curling back on itself inside the cabin. Fresh air mode is the same outline with an arrow entering from outside. If your dash has one button that toggles between them, the illuminated state is almost always recirculation.

Why does recirculated air make you drowsy on a long drive?

The proposed mechanism is carbon dioxide. You exhale roughly 40,000 ppm of CO2 with every breath. In fresh air mode that gets flushed continuously. In recirculation the flushing slows way down, so CO2 accumulates until the rate you produce it matches the rate the car leaks it out. In a car full of people, that balance point is high.

The part everyone gets wrong is the word "sealed." A car in full recirculation is not sealed. Hudda and Fruin measured air exchange rates of 3 to 23 air changes per hour in recirculation mode versus 45 to 104 in outside air mode, driving at 15 to 60 km/h. Body seams, door seals and pressure relief vents keep leaking regardless of what the flap is doing. That is why cabin CO2 plateaus instead of climbing forever, and it is why nobody has ever suffocated in a running car.

So the effect is real and measurable. The size of it is where this gets complicated.

How high does CO2 actually get inside a car cabin?

Peer-reviewed measurements put a sealed-up cabin somewhere between 2,000 and 3,000 ppm in normal use, with worst-case figures near 6,000 ppm for a full car over a longer drive. Outdoor air sits around 400 to 420 ppm. Occupant count matters more than anything else, including how long you have been driving.

Read the two NIOSH rows together, because that is the useful finding in this whole article. Going from 100 percent recirculation to 75 percent recirculation with a 25 percent fresh air bleed dropped the cabin from about 3,000 ppm to about 1,000 ppm. You do not have to choose between cold air and clean air. A small amount of outside air does most of the work.

Rate of rise depends almost entirely on how many people are breathing. Wei found that holding a cabin under 1,000 ppm required cutting recirculation from 90 percent with one passenger down to 18 percent with five. Mathur's SAE field testing found levels climbing past reference thresholds within the first five minutes of recirculation with a single occupant.

Is the science on cabin CO2 and driving actually settled?

No, and any article that tells you otherwise is selling something. The measurements above are solid. What those concentrations do to a driver is genuinely contested, and the most-cited study in this space has already failed to replicate. Here is the honest version, because you deserve it before you go rearranging your build.

The famous one is Satish et al. 2012, from Lawrence Berkeley National Laboratory, published in Environmental Health Perspectives. Twenty-two subjects were tested at 600, 1,000 and 2,500 ppm. Six of nine decision-making measures declined at 1,000 ppm, and seven of nine at 2,500 ppm. The authors themselves wrote that the strength of the effects at 2,500 ppm "is so large for some metrics as to almost defy credibility," and called for replication before anyone drew conclusions.

The replication came in 2019, published in npj Microgravity, testing 600, 1,200, 2,500 and 5,000 ppm on 22 astronaut-like subjects. It did not reproduce the result. At 2,500 and 5,000 ppm, performance was similar to or better than baseline. Usha Satish is a co-author on that paper too, so this is not a case of rival labs taking shots at each other.

The one study that put actual drivers in a simulator, Shimazaki et al. 2025, tested eight taxi drivers at under 500 ppm versus 5,000 ppm. Driving errors did increase significantly: wobble, contact and wheel departure all got worse. But reaction time got faster at high CO2, not slower, and working memory was unaffected. On sleepiness specifically, Inada et al. 2024 found shortened sleep onset latency at 4,851 ppm compared to a 1,102 ppm control, with no measured cognitive effect.

For scale, the OSHA permissible exposure limit for CO2 is 5,000 ppm averaged over eight hours. Typical recirculation levels sit well below that. And there is no crash data behind any of this. NHTSA's Crash Investigation Sampling System does not record HVAC state, and neither does Arizona's crash reporting. Anyone claiming investigators check the recirculation setting is making it up.

The fair summary: stale cabin air is real, plausibly makes you feel duller and stuffier on a long haul, and costs you nothing to fix. Treat it as comfort and good practice, not as an emergency.

Should you use recirculation on the highway?

Use it, but do not leave it on for the entire drive. The practical rule that matches the measurements is to run a fresh air bleed continuously if your car supports partial recirculation, or to toggle to fresh air for a few minutes every 30 to 60 minutes if it does not. With three or more people in the car, do it closer to every 30.

Occupancy is the variable that should drive your decision. Alone in the car on a two-hour highway run, recirculation is close to harmless. Four people in a coupe with the windows up is the scenario where the numbers get interesting, and it is also the scenario where the most people are relying on you to stay sharp.

If your car has automatic climate control, check whether it auto-cycles out of full recirculation. Many do, though the documented reason in manufacturer patents is windshield fogging and comfort, not CO2. Do not assume the car is managing air quality for you.

When is recirculation the right setting?

Recirculation is not the villain. It is the correct setting more often than the viral posts admit, and in a classic with a marginal A/C system it is frequently the only way to get genuinely cold.

• Stop and go traffic. You are sitting in the exhaust of the car in front of you. Close the door.
• Tunnels and parking structures. Same reason, worse concentrations.
• Behind a diesel, or anything visibly smoking. Recirculate until you have passed it.
• Wildfire smoke and dust storms. Arizona gets both. Recirculation plus a good cabin filter is the right answer.
• Initial cooldown. Pulling 140 degree air out of a car that has been sitting in the sun is much faster on recirculation.
• Any classic where the A/C is barely keeping up. This is most of them. Get it cold, then bleed fresh air back in.

The message is not to avoid recirculation. It is to recirculate deliberately instead of permanently.

How classic cars breathed before air conditioning

Before A/C was standard, ventilation was the whole design. GM introduced No-Draft Ventilation in 1933, later branded Ventipanes, the pivoting vent window in the front door corner. The patent was filed in November 1932. Those little triangular windows are not decoration. They were the climate control system.

The other half was flow-through ventilation. High pressure air enters at the cowl, where the base of the windshield meets the hood, collects in the cowl plenum, and feeds the dash vents and kick panel vents. Meanwhile the air moving past the sides of the car creates low pressure, which pulls cabin air out through extractor vents behind the rear seat or in the door jambs. The car was designed to breathe front to back at speed, using nothing but a pressure differential.

GM branded its version Astro Ventilation starting in 1968, right as vent windows began disappearing. The 1968 system used passive flaps and 1969 added a vacuum motor. Here is the part that matters: only non-A/C cars got Astro Ventilation. Factory air conditioned cars received block-off plates over those openings. GM traded flow-through ventilation for air conditioning almost sixty years ago, and the aftermarket has been repeating that trade ever since.

Does aftermarket classic car air conditioning have a fresh air mode?

Often, no. This is the finding that surprised us, and it is documented in the manufacturers' own installation instructions rather than in any review or forum rumor. Many retrofit A/C systems are recirculation-only by design, and their manuals explicitly instruct the installer to permanently block off the factory fresh air inlet at the cowl.

Two things to be fair about. First, kits vary enormously by application and manufacturers revise them, so confirm with your supplier for your specific car rather than taking a table on the internet as gospel. Second, a missing fresh air mode is not a knock on cooling performance. These systems exist to make a hot classic genuinely cold in Phoenix traffic, and the good ones do that extremely well. It just means the fresh air path becomes your problem to solve rather than theirs.

Why a finished restomod breathes worse than the original car did

An original, unrestored classic leaks like a colander. Dried out weatherstrip, gapped door seals, a rusty cowl, holes in the firewall where somebody ran a stereo wire in 1987. Research on vehicle air exchange rates consistently finds that older, higher-mileage cars leak more than new ones. CO2 never gets a chance to accumulate. The tradeoff is that everything else comes in with it: exhaust, fuel vapor, road dust and brake dust, unfiltered.

A finished build inverts that completely. New weatherstrip seals. Sound deadening closes up the floor and firewall. Grommets get installed properly. One-piece door glass deletes the vent window. Bonded flush-mount glass eliminates the rubber gasket that was quietly passing air around the windshield for fifty years. Every one of those is an upgrade, and together they produce a cabin as tight as a modern car.

Then a recirculation-only A/C unit goes in and the cowl inlet gets capped per the instructions.

The result is a car that is tighter than the 1969 version of itself and tighter than a 2026 sedan, which at minimum still has a fresh air blend door. That is the restomod paradox. Nobody did anything wrong. Every individual decision was correct. The fresh air path just quietly disappeared somewhere between the weatherstrip and the evaporator, and nobody noticed because there is no warning light for it.

We say this as the company selling you a lot of those sealing parts. Flush-mount glass is a real upgrade for water intrusion, wind noise and A/C performance, and we will keep telling you to buy it. We would just rather you knew what a genuinely tight cabin changes. The same goes for a C10 vent window delete. It looks better, seals better, and it removes an air path. If you delete it, replace what it was doing.

Is that drowsy feeling CO2, or is it carbon monoxide?

In a classic, this is the question that actually matters, and the answer is not the one the viral posts are chasing. CO2 makes you dull. Carbon monoxide kills people. If you feel headachy, drowsy or nauseated in an older car with a header leak or an unsealed firewall, stop diagnosing your HVAC and start looking for exhaust.

The CDC reports more than 400 unintentional non-fire carbon monoxide deaths in the US each year, more than 100,000 emergency department visits and more than 14,000 hospitalizations, and states plainly that a small leak in the exhaust system can build up CO inside the car. The NTSB's guidance lists the entry points as cracked mufflers, leaking pipes, unplugged firewall holes and degraded door and window seals. That is a precise description of a hot rod with open headers and a firewall full of unsealed pass-throughs.

Note the 200 ppm row. Loss of judgment, at a concentration a leaky classic can absolutely produce. That is a far better documented driver-impairment story than anything in the CO2 literature, and almost nobody writing about the recirculation button mentions it.

Buy a battery powered electrochemical CO monitor. They are about thirty dollars, they fit in a cupholder or glovebox, and for an old car with a modified exhaust they are the single highest-value safety item you can add. Do not rely on smell. Carbon monoxide has none.

Where do exhaust fumes get into a classic car cabin?

Almost always through a hole somebody forgot about. The usual suspects, roughly in order of how often we see them: unsealed firewall pass-throughs for wiring, cables and A/C lines; missing or hardened grommets; a rusted or poorly sealed cowl plenum; trunk and deck lid seals that no longer seal; floor pan repairs that were never fully welded or seam sealed; and tailgate or hatch seals on trucks, wagons and Blazers.

The area directly behind a moving car is a low pressure zone. That suction is why cracking only a rear window or driving with a tailgate open can pull exhaust forward into the cabin. The CDC states it directly for tailgates: if only the tailgate is open, CO from the exhaust will be pulled into the vehicle. Open a front window or the vents too, so air is actually moving through.

The cowl deserves special attention, because it is both the fresh air inlet and the most common rust location on these cars. If the cowl plenum is rotten, you are pulling air through rust and whatever is under the hood. Our guide on telling a windshield leak from a cowl leak covers how to test yours, and the monsoon-proofing checklist covers sealing it properly.

How do you get fresh air into a restomod without killing the A/C?

The goal is a controlled, closeable fresh air path, not a permanent hole. An open hole will destroy your cooling performance in Phoenix and fog your glass in winter. A door that seals when closed gives you both.

1. Find out what your A/C system actually has. Call the manufacturer with your kit number. Ask specifically whether a recirculation and fresh air door assembly is available for your application. On some Vintage Air Gen IV and Gen V setups it is an available part, and most people never knew to ask.
2. Keep the kick panel and astro vents functional, but make them seal. Builders universally report that leaking vent doors kill cooling capacity. The fix is a door that closes properly, not a vent siliconed permanently shut.
3. If your installer capped the cowl, consider restoring a controlled inlet. Plenty of builders have added a small ducted, closeable fresh air feed back into the box. Do it with a door, not an open pipe.
4. Fix exhaust leaks and seal the firewall before you chase A/C performance. Every unsealed pass-through is both an air leak and a CO path. Seal them and your A/C works better at the same time.
5. Buy the CO monitor. Thirty dollars. Do this one first, honestly.
6. Then just crack a window. For a few minutes every half hour on a long drive, in a car with no fresh air mode at all, this is the entire fix. It is not sophisticated and it works.

If you are already deep in a dash or interior job, that is the moment to sort all of this out, because the HVAC box is accessible exactly once. Our dash planning guide and the dash overlay versus full restoration breakdown both cover the right sequencing. If you are replacing vents while you are in there, the billet A/C vents we stock are a straightforward upgrade over cracked factory louvers.

Quick answers

Does recirculation save gas? Marginally, in hot weather. The compressor cycles less because it is cooling already-cooled air. The effect is small and mostly relevant on long hot drives.

Does recirculation reduce the oxygen in your car? Not meaningfully. Oxygen drops by a fraction of a percent while CO2 rises. The CO2 is what changes measurably, not oxygen depletion.

Why do my windows fog up when recirculation is on? Because you are recycling your own breath, which is humid. Switch to fresh air and turn on the A/C compressor, which dehumidifies. This is also why many modern cars auto-defeat recirculation in defrost mode.

Can leaving recirculation on damage my A/C system? No. It reduces the load on the system. The downsides are cabin air quality and humidity, not mechanical wear.

Do I have to block off the cowl vent when installing aftermarket A/C? Many installation manuals instruct you to, because an open inlet lets hot outside air into the box and hurts cooling. A sealing door is the better solution than a permanent block-off if your kit supports one.

Is CO2 in a car ever actually dangerous? Measured cabin levels stay well under the 5,000 ppm OSHA eight-hour exposure limit. The realistic effect is feeling stale and dull, not a medical emergency. Carbon monoxide from an exhaust leak is the genuine hazard in an older car.

Does flush-mount glass make this worse? It makes the cabin tighter, which is the point, and that is a net win for noise, water and A/C performance. It just means fresh air needs to come from somewhere deliberate.

Shop Fesler flush-mount glass or browse billet A/C vents. Building a car in Arizona and not sure what to seal first? Talk to us.