Indoor air quality is the area where cheap sensors have changed the most in a decade, and where the numbers are most often misread. A monitor that says “poor” in red is not very useful. A monitor that says 1,400 parts per million of carbon dioxide, and a person who knows what that means, is genuinely useful.
We spend the large majority of our lives indoors, and indoor air is frequently worse than outdoor air. Here is what is worth measuring, what the figures mean, and what to actually do about them.
Carbon dioxide: the ventilation proxy
Carbon dioxide is not itself very harmful at indoor concentrations. Its value is as a proxy for ventilation: people exhale it constantly, so if it is accumulating, fresh air is not arriving fast enough — and everything else people exhale, including airborne viruses, is accumulating alongside it.
Rough guide, in parts per million:
- 400–450: outdoor background. You cannot go below this indoors.
- Under 800: well ventilated.
- 800–1,200: acceptable, ventilation is being outpaced.
- 1,200–1,500: stuffy. Measurable effects on concentration and decision-making appear in this range in study after study.
- Above 1,500: open a window. A closed bedroom with two adults routinely exceeds 2,000 by morning.
Insist on a true NDIR sensor — non-dispersive infrared, which measures CO₂ directly by how it absorbs infrared light. Many cheap monitors instead estimate CO₂ from a volatile-compound sensor and label it “eCO₂” or “CO₂ equivalent”. Those readings respond to cooking, cleaning products and aftershave, and they are not a ventilation measurement. The distinction is the single most important thing to check when buying.
Fine particles: the one with the clearest health evidence
PM2.5 means particulate matter under 2.5 micrometres across — small enough to reach deep into the lungs and into the bloodstream. The epidemiological evidence linking long-term particulate exposure to cardiovascular and respiratory harm is among the strongest in environmental health, which is why the World Health Organization has progressively tightened its guidance.
Indoors, the dominant sources are usually not outdoors. Cooking — particularly frying, grilling and anything involving hot oil — produces spikes that dwarf typical outdoor levels. Candles, open fires, wood burners, incense and smoking all produce large amounts. Vacuuming without a good filter redistributes it.
The WHO guideline for annual average exposure is 5 µg/m³, with a 24-hour guideline of 15. Domestic monitors are not laboratory instruments and their absolute accuracy is limited, but they are good at showing you relative change, which is what changes behaviour. Watching the number go from 8 to 300 while you sear a steak, and back down within twenty minutes of the extractor running, teaches you more than any guideline.
Volatile organic compounds, honestly
VOCs are gases emitted by solids and liquids — paint, adhesives, new furniture, cleaning products, air fresheners, printers, and your own body.
Be sceptical of what domestic VOC sensors tell you. They report a single combined figure, cannot distinguish a harmful compound from a harmless one, and drift over time. A VOC sensor cannot tell formaldehyde from the orange you just peeled.
Used sensibly they are still informative: a persistent elevated baseline in a newly decorated or newly furnished room is worth acting on with ventilation, and a spike whenever a particular product is used is a reasonable prompt to use less of it. Treat the number as a relative signal, never as a safety threshold.
Radon: the one you cannot sense and must test for
Radon is a naturally occurring radioactive gas that seeps from certain rocks and soils into buildings, accumulating in basements and ground floors. It is the second leading cause of lung cancer after smoking in many countries, and it is completely undetectable without a test.
Crucially, risk is geological and local — two houses in the same town can differ substantially. Most countries publish a radon risk map, and if you are in an affected area a long-term test is the only way to know. Continuous radon monitors now exist at domestic prices and are worth it in affected regions; elsewhere a one-off long-term test is proportionate.
Mitigation is well established and effective: sealing, sub-floor ventilation, or a sump and fan. This is not a problem you automate away, but ventilation automation can help, and knowing is the whole point.
Humidity, which affects everything else
Relative humidity has a comfortable band of roughly 40 to 60 per cent, and both ends of the scale cause problems.
Too dry — below about 30 per cent, common in winter with strong heating — irritates airways and eyes and can worsen asthma. Too damp — sustained above about 65 per cent — allows dust mites and mould to thrive, which is a well-documented driver of respiratory symptoms and allergy.
Humidity is also the cheapest thing to measure and among the most actionable, because the fix is usually ventilation or a simple behaviour change rather than equipment.
What not to bother with
Combined “air quality index” scores that reduce everything to one number. They obscure which parameter is out of range, which is the only actionable information.
Monitors without a true CO₂ sensor, if ventilation is what you care about.
Anything claiming to measure allergens, bacteria or viruses directly at a domestic price. It cannot.
Ozone-generating or ionising purifiers. Ozone is a respiratory irritant and several health bodies advise against generating it indoors deliberately.
Acting on the numbers, in order of effectiveness
Ventilate. The most effective intervention for CO₂, VOCs and humidity, and it costs nothing. Opening windows on opposite sides of a home for even five minutes clears a remarkable amount, far more efficiently than one window open all day.
Extract at source. A cooker hood that vents outside, used before you start cooking rather than after the smoke alarm reacts, removes cooking particles at the point they are generated. A recirculating hood with a grease filter does very little for fine particles.
Filter what you cannot ventilate away. A purifier with a genuine HEPA filter is effective for particulates, and only for particulates. Size it to the room by air changes per hour, and note that a large unit on a low setting is quieter and more effective than a small one at maximum — and a purifier you turn off because it is noisy filters nothing.
Then automate. CO₂ above 1,000 in a room, run the ventilation or notify. Humidity above 65 in a bathroom, run the extractor until it drops. PM2.5 spike, run the purifier and the hood. Automation is the last step because it multiplies good decisions and cannot substitute for them.
A sensible starting kit
- One good monitor with a true NDIR CO₂ sensor plus PM2.5, in the room where you spend most waking time.
- A cheap temperature and humidity sensor in every room. These cost very little and reveal damp problems, cold spots and ventilation failures.
- A radon test if you are in an affected area.
- Move the good monitor to the main bedroom for a week. Overnight CO₂ is the reading that surprises people most, and it is the easiest to fix.
- Then buy equipment, once you know which parameter is actually out of range in your home rather than guessing.
Step four is the one we would push hardest. Almost everyone who measures their bedroom overnight changes something the following week.