There are two types of CO2 monitors on the market. One measures CO2. One guesses.
The packaging looks identical. The price difference can be 5x. And if you buy the wrong one, you will never know your air is making you dumb — because the number on the screen is fiction.
I've designed air quality products. I've read the sensor spec sheets, sourced the components, and shipped devices with both sensor types inside. Here's the actual difference.
NDIR stands for Non-Dispersive Infrared.
CO2 molecules absorb infrared light at a very specific wavelength: 4.26 micrometers. An NDIR sensor shines an IR light source through a sample chamber. A detector on the other side measures how much of that specific wavelength gets absorbed. More absorption = more CO2 molecules in the air.
That's it. You are measuring CO2 directly. Physics, not inference.
Good NDIR sensors also compensate for temperature and pressure drift, and run automatic baseline calibration (ABC) — the assumption being that the sensor sees fresh outdoor air (~420 ppm) at least once every few days.
This is the layer most reviews skip.
Consumer monitors don't manufacture the CO2 sensor inside. They buy it from a component supplier and build around it. Knowing who made the sensor tells you more than the brand name on the box.
The reputable NDIR manufacturers:
Senseair — Swedish company. Their S8, LP8, and Sunrise modules are in a huge share of professional and prosumer devices. If a monitor lists a Senseair sensor, that's a good sign.
Sensirion — Swiss. The SCD40 and SCD41 are widely used in mid-range monitors and DIY builds. Solid accuracy, well-documented. Sensirion also makes the SGP series — but those are MOX sensors (more on that below). Same company, completely different technology. Don't confuse them.
Cubic — Chinese manufacturer, CM1106 and CM1107 modules. Common in budget-to-mid NDIR devices. Functional, less premium than Senseair or Sensirion.
Amphenol/Telaire — More commercial and HVAC-oriented. Well-regarded in building automation, less common in consumer devices.
Vaisala — Finnish, industrial grade. Their CO2 transmitters show up in labs and building management systems, not on Amazon. The gold standard at the high end.
If a monitor's product page doesn't name the sensor manufacturer — ask. Or assume it's not one of the above.
eCO2 stands for "estimated CO2."
That word — estimated — is doing a lot of work.
eCO2 sensors are MOX sensors: Metal Oxide sensors. They detect total volatile organic compounds (VOCs) in the air. The idea is that humans exhale both CO2 and VOCs, so if VOC levels rise, CO2 probably did too.
The problem: everything else also produces VOCs. Cleaning products. Furniture off-gassing. Cooking. A new rug. Someone using hand sanitizer across the room. The sensor can't tell the difference between a crowded meeting room and a freshly cleaned empty one.
So the eCO2 number is a statistical guess. The sensor runs a model — "here's a VOC reading, here's what CO2 tends to look like alongside that" — and outputs a number that looks like a CO2 reading.
It is not a CO2 reading.
You cannot calibrate it into accuracy. The fundamental measurement is wrong.
The main eCO2 sensor suppliers: Bosch (BME680, BME688), ScioSense/ams (CCS811, ENS160), and — again — Sensirion's SGP30/40 line. Good companies. Wrong tool for this job.
Cost.
A decent NDIR sensor module costs a few dollars at volume. A MOX sensor costs cents.
For a manufacturer optimizing for Amazon price point, the math is obvious. Slap "CO2 monitor" on the box, put an eCO2 sensor inside, price it at $29.99. Most buyers will never know.
The spec sheet tells the truth if you read it. Look for "NDIR" in the sensor description. If it says "eCO2," "estimated," "equivalent CO2," or just lists VOC with a CO2 output — that's a MOX sensor, not a CO2 sensor.
A monitor that makes up numbers is worse than no monitor. It can show 800 ppm when the room is at 1,600 ppm. You'll sit there confident while your thinking slows down.
If you are buying a CO2 monitor, the only question that matters: does it have an NDIR sensor?
If yes: it measures CO2. Buy it.
If no: it guesses. Don't.
Is eCO2 ever accurate?
Occasionally, in very controlled conditions with stable backgrounds, eCO2 readings can loosely track real CO2. But they can't be calibrated to a reference, and any VOC source in the room will throw them off. For anything you want to act on, no.
How do I know what sensor my monitor has?
Look for "NDIR" on the product page or in the spec sheet. If the listing says "eCO2," "estimated CO2," "equivalent CO2," or just mentions VOC detection alongside a CO2 output — it's a MOX sensor. If the sensor type isn't mentioned at all, assume the worst and contact the manufacturer before buying.
Can a MOX sensor be used alongside an NDIR sensor?
Yes, and some higher-end monitors do exactly this. A MOX sensor measuring VOCs is genuinely useful data — it just shouldn't be labelled as CO2. If a device has both an NDIR sensor for CO2 and a separate MOX sensor for VOCs, that's a feature. The problem is when there's only a MOX sensor and it's sold as a CO2 monitor.