A CO₂ monitor accuracy specification such as ±(50 ppm + 3% of reading) means the true value is within 50 ppm plus 3 percent of the displayed number. At 1,000 ppm that is ±80 ppm. That is more than enough for deciding when to ventilate. In practice, placement, warm-up and baseline drift cause far larger errors than the sensor specification.
Reading the specification
The fixed part (30 to 50 ppm) covers noise and offset; the percentage part covers scaling error. Combine them at the level you care about:
| Spec | At 600 ppm | At 1,000 ppm | At 2,000 ppm |
|---|---|---|---|
| ±(30 ppm + 3%) | ±48 ppm | ±60 ppm | ±90 ppm |
| ±(50 ppm + 3%) | ±68 ppm | ±80 ppm | ±110 ppm |
| ±(50 ppm + 5%) | ±80 ppm | ±100 ppm | ±150 ppm |
Manufacturers state the figure at a reference temperature (usually 25 °C) after calibration. Some also quote a separate drift figure per year.
What accuracy you actually need
The decision thresholds are 800, 1,000 and 1,500 ppm, spaced hundreds of ppm apart. A ±100 ppm window does not change the decision. Tighter accuracy matters only for research, for verifying an HVAC system against a standard, or for comparing readings across many monitors.
The errors that matter more
Placement
A person’s exhaled breath is around 40,000 ppm. A monitor on the desk in front of you can read 300 to 800 ppm above the room. Keep it a meter or more from the nearest person and away from vents and windows.
Warm-up
NDIR lamps and photoacoustic sensors need a minute or two to stabilize after power-on. Ignore the first few readings.
Baseline drift and calibration
Optical paths drift over months. Automatic baseline correction assumes the room reaches outdoor levels (about 420 ppm) at some point each week and re-zeros to that. In a room that never empties, or a monitor that lives in a closed bedroom, the baseline creeps and every reading is offset. A manual fresh-air calibration fixes it in ten minutes. See how to calibrate a CO₂ monitor.
Pressure and altitude
Absorption depends on the number of molecules in the light path, which falls with altitude. Monitors with a pressure sensor compensate; others read about 1 percent low per 100 m of elevation unless configured.
Wrong technology
None of the above applies to eCO₂ estimators, which can be off by a thousand ppm regardless of specification. See NDIR vs metal-oxide sensors.
Checking a monitor yourself
- Put it outdoors or by a wide-open window, out of direct sun, for 15 minutes. It should read roughly 400 to 450 ppm.
- Bring it into an occupied room a meter from people. It should climb steadily over the next half hour.
- Breathe gently toward it from arm’s length; it should spike and then recover within a few minutes.
A monitor that passes those three checks is accurate enough for any household or office decision. The best CO₂ monitors guide lists the stated accuracy of each pick.
Frequently Asked Questions
How accurate is a typical consumer CO₂ monitor?
NDIR consumer monitors usually state ±(30 to 50 ppm) plus 3 to 5 percent of the reading. At 1,000 ppm that is a window of roughly ±60 to ±100 ppm.
Do two monitors in the same room read the same?
Usually within 50 to 100 ppm once both are warmed up and calibrated. Larger differences almost always come from placement or a drifted baseline on one unit.
Does humidity affect CO₂ readings?
Very little for NDIR sensors at normal indoor humidity. Condensation inside the sensor does, so avoid bathrooms and cold windowsills.