You plug it in, the fan comes up, and within an hour the room feels better. That feeling is the sound of moving air, not a measurement.
Air cleaners are the easiest thing to buy and the hardest thing to buy well. The Environmental Protection Agency puts them third in a fixed order of operations: get rid of whatever is producing the pollution, then bring in clean outdoor air, then filter what is left. EPA describes filtration as a supplement to those first two steps, not a stand-in for either. Its plain warning about both portable units and furnace filters is that they “cannot remove all pollutants from the air.”
None of that makes filtration pointless. It makes it the third move. And a third move pays off only if you know what the machine is doing.
This article covers the devices and the filters. Whole-home ventilation is a separate subject, and so is the stove.
Filtration is the third step, and the order is the point
EPA lists three basic ways to improve indoor air: source control, improved ventilation, and air cleaners or filtration. The agency says the most effective approach is usually to eliminate individual sources of pollution or reduce what they emit. It also notes that source control is often cheaper than increasing ventilation, which drives up heating and cooling costs.
Picture that in an actual room. If a scented plug-in is putting chemicals into your bedroom air, a filter cannot out-run it. The filter cleans while the source refills. Unplug the source and the problem is gone, permanently, for free.
Filtration earns its place when the source will not go away. Pollen that rides in on clothes. Wildfire smoke already inside. Dust from a street you cannot move. Dander from a dog you are not giving up. A correctly chosen filter helps with all of those.
The two kinds of filters in an American home
EPA splits residential filtration into two categories. Portable air cleaners, sometimes labeled air purifiers or air sanitizers, filter one room or one open area. Furnace or central HVAC filters sit in the ductwork and filter air throughout the house.
The central filter has an advantage portable units cannot match: it treats every room the system serves. It also has a limitation almost nobody accounts for at purchase. According to EPA, central systems run only when heating or cooling is called for, so air cleaning is typically limited to less than 25 percent of the 8,760 hours in a year. No air moving through the duct means no air going through the filter.
Most thermostats have a fan setting that circulates air without heating or cooling. Running it longer filters more air. It also costs more in electricity, and extended fan operation can work against humidity control in cooling season, because moisture that condensed on the coil gets blown back into the house.
Reading the ratings without getting lost
Three numbers show up on packaging, and each one answers a different question. None of them appears on EPA’s general air cleaners page, so each is worth tracing back to the document that defines it.
MERV, for the filter in your furnace
MERV stands for Minimum Efficiency Reporting Values. EPA describes it as a measure of a filter’s ability to capture particles between 0.3 and 10 microns, using a test method developed by the American Society of Heating, Refrigerating and Air-Conditioning Engineers. EPA’s own table runs from MERV 1 to MERV 16, and the agency notes that HEPA filters sit outside that scale entirely because they are not MERV-rated.
EPA’s technical summary on residential air cleaners is more specific about what the numbers buy you. The rating comes from ASHRAE Standard 52.2. Generally, a higher rating means higher removal efficiency for at least one particle size range, and MERV 13 and above require at least 50 percent removal efficiency for particles between 0.3 and 1 micron. That band covers much of the fine particle range, which is why MERV 13 keeps coming up as a floor rather than a luxury.
HEPA, which is a performance claim and not a brand
HEPA stands for high efficiency particulate air, and EPA points to the Department of Energy as the body that formally defines it. A HEPA filter can theoretically remove at least 99.97 percent of airborne particles at 0.3 microns.
The counterintuitive part is worth a sentence. EPA explains that 0.3 microns is the most penetrating particle size, the hardest size for a filter to catch. Particles both larger and smaller are captured at even higher efficiency. So a HEPA rating is not a floor at the small end. It is a floor at the worst case, which is a stronger claim than it sounds.
Then the correction. EPA’s technical summary states that there is no widely accepted definition of HEPA performance in consumer products, and that residential filters sold as HEPA are generally equivalent to MERV 16. The word on the box is a claim, not a certification. Look for the stated efficiency and the particle size next to it.
Clean air delivery rate, for portable units
Clean air delivery rate, or CADR, tells you how much filtered air a portable unit delivers. EPA’s Residential Air Cleaners: A Technical Summary explains that it combines removal efficiency with airflow, and that units are tested under procedures set by the Association of Home Appliance Manufacturers. The same document adds that the resulting room-size rating is meant to deliver roughly an 80 percent reduction in particle levels.
Two limits matter. CADR is a particle-only number, so it says nothing about gases. And EPA states the rating is typically measured at the air cleaner’s highest speed. Hold on to that second point. It explains a very common disappointment.
Why a higher MERV number can work against you
More filtration sounds like more protection. In a ducted system it is not that simple.
A furnace filter is a restriction the blower has to pull air through. EPA’s technical summary says filters rated MERV 14 to 16 typically have higher average resistance to airflow. Any higher-efficiency filter has to be checked against the existing ducted system so airflow is not impeded. High pressure drops on some fibrous media filters can negatively affect HVAC equipment, and EPA notes a retrofit may need a trained professional to modify the system.
Picture a drinking straw packed with cotton. You are still drinking, just much less, and working much harder. A blower pulling against too much resistance moves less air, so less air gets filtered, rooms heat and cool unevenly, and the equipment runs hotter than it was designed to.
EPA’s practical answer sets a floor and then a limit: choose at least MERV 13, or the highest rating your system’s fan and filter slot can accommodate, and ask an HVAC professional which filter works best with your equipment. EPA notes that residential filters come in thicknesses from one to five inches, so a deeper filter slot is worth asking about during that conversation.
Fit matters as much as rating. EPA’s technical summary warns that air flowing around a poorly installed filter reduces its effectiveness, and that duct-mounted performance drops substantially with leakage around the frame. A filter should slide in snugly without being bent or crushed. Air takes the easiest path, and a quarter-inch gap is easier than any filter media on the market.
Sizing, running time and the filter you forgot
Size it to the room, then run it quietly
EPA’s guide gives a simple sizing method: multiply the room’s length by its width, then compare that square footage against the maximum room size on the product packaging. Its estimate chart assumes an eight-foot ceiling and lists minimum clean air delivery rates of 65 cubic feet per minute for 100 square feet, 195 for 300 square feet, and 325 for 500 square feet. Higher ceilings mean choosing a unit rated for a larger area. An open floor plan means counting the whole connected space, not the part you think of as the room.
Now put sizing together with that detail about how CADR is tested. The rating is measured at the highest speed, and EPA states plainly that an air cleaner filters less air at a lower speed. So the number printed on a box is a promise about a setting many people never use.
Here is the scenario that plays out in bedrooms across the country. Somebody buys a unit rated right at the edge for their room, runs it on the quietest setting all night because the high setting sounds like a hair dryer, and concludes after a month that air cleaners do not work. The unit was never undersized on paper. It was undersized in practice from the first night.
The fix is to buy a unit rated well above the room, then run it on low or medium. A larger machine loafing moves more clean air than a small machine straining, and it does it more quietly.
Placement follows the same logic. A filter only does something to air that passes through it. EPA advises putting units in the rooms where you spend the most time, keeping airflow unobstructed and away from curtains, and using them on the floor or a tabletop as designed. If the airflow creates an uncomfortable draft, point it away from you rather than turning the unit down. EPA’s guidance for do-it-yourself air cleaners is to run them the entire time the space is occupied.
A neglected filter is a worse filter
Maintenance decides whether the purchase was worth anything a year later. EPA is direct about it: all filters need regular replacement, and a dirty, overloaded filter will not work well. For furnace and central HVAC filters, the guide notes that manufacturers typically recommend replacement every 60 to 90 days, and sooner if a filter looks heavily soiled. Portable units have their own schedules, and many have a pre-filter that needs vacuuming far more often than the main filter needs replacing.
A five-minute check you can do this week:
- Find the filter. Write down the dimensions printed on the frame so you are not guessing at the store.
- Hold it up to a light. If light does not pass through it, you are past due.
- Look at how it sits in the slot. Gaps on any side mean air is going around it.
- Write today’s date on the frame edge in marker.
- Set a reminder 60 days out, then adjust the interval based on what the next filter looks like.
Households with pets, a smoker, construction nearby or a wildfire season will go through filters faster than the label suggests. The label is a starting estimate, not a measurement of your house.
What a filter cannot do
This is the limit that sends the most people down the wrong road.
EPA’s guide states that most filters are designed to filter either particles or gases, and the technical summary is blunter: devices built only to remove particles are incapable of controlling gases and some odors. A particle filter catches particles. Gas molecules are orders of magnitude smaller than anything it is built to trap, and they pass straight through.
Some units add activated carbon or another sorbent for gases. EPA notes that those media have a finite capacity and need a large amount of material to do much, that there is no widely used consumer rating system for gas removal, and that carbon performs poorly against some compounds. A thin carbon mat behind a HEPA filter is mostly reassurance. EPA also reports that research has not shown plasma, photocatalytic oxidation or ultraviolet light to remove gases effectively in residential units.
Mold is the other common mismatch. Filters do not fix mold. The moisture has to be eliminated and the mold cleaned up. A filter may capture some mold-generated particles and reduce some odors, which is not the same as solving anything. EPA also says it does not currently recommend air cleaners for reducing radon.
The gas limitation is exactly why a filter is the wrong answer to a stove problem. Nitrogen dioxide and carbon monoxide from a gas burner are gases. A particle filter will not touch them, however many stars the box has. The answer at the stove is exhaust that goes outside, which belongs to the cooking article rather than this one.
Ozone generators sold as air cleaners
Some devices are marketed as air cleaners and work by deliberately producing ozone. EPA’s position on them is not soft, and this article will not soften it.
EPA states that at concentrations that do not exceed public health standards, ozone has little potential to remove indoor air contaminants, and that it is generally ineffective at controlling indoor air pollution. It is not effective at removing many odor-causing chemicals either. Body odor may be masked by the smell of ozone without being removed.
On health, EPA says inhaled ozone can damage the lungs. It can cause chest pain, coughing, shortness of breath and throat irritation, and it can worsen asthma. The agency reports that these generators can drive indoor ozone above public health limits even when people follow the manufacturer’s instructions.
Ozone also reacts with other chemicals already in a house, and EPA says those reactions can form harmful or irritating by-products. In one laboratory test, ozone reacting with chemicals from new carpet produced aldehydes while total airborne organic chemicals went up. The device sold to clean the air added to what was in it.
Some sellers claim federal approval. On that, EPA’s technical summary states: “No federal agency has approved the use in occupied spaces of air cleaners that intentionally emit ozone.”
Ozone can also be an unintended by-product. EPA names electrostatic precipitators, ionizers, ultraviolet lights without adequate lamp coatings and plasma cleaners as device types that may emit some. The California Air Resources Board publishes a list of certified air cleaning devices and the Association of Home Appliance Manufacturers publishes a verified list of room air cleaners, and both are free to check. One more label to distrust: EPA points out that ENERGY STAR on an air cleaner indicates energy efficiency only. It says nothing about how well the unit cleans air.
What the research shows, and what to do with it
Put the evidence in tiers, because the tiers are different.
That these devices reduce indoor particle concentrations is established. In a randomized trial published in the journal Indoor Air in 2021, Riederer and colleagues placed two portable HEPA cleaners in the homes of children aged 6 to 12 with asthma in Washington State’s Yakima Valley. Across 71 households, homes that received cleaners had 60 percent lower fine particle concentrations in the sleeping area and 42 percent lower in the living area at one-year follow-up. Ammonia, a gas, did not drop. The particle-versus-gas distinction showed up in the data of a real trial.
Whether people feel better is a weaker claim. Drieling and colleagues reported health results from that same cohort in Environmental Health in 2022. The primary measure of asthma control did not differ significantly between groups. Several secondary measures favored the air cleaner group, and the authors cautioned that the study was unblinded and many outcomes self-reported.
A 2026 systematic review and meta-analysis in the Indian Journal of Occupational and Environmental Medicine by Mudgal and colleagues pooled 27 randomized controlled trials of personal-level measures including air purifiers. Indoor fine and coarse particle levels fell substantially. Lung function, respiratory symptoms and blood pressure showed no significant improvement. EPA’s own guide describes the literature on portable HEPA units as showing small improvements in cardiovascular and respiratory measures, with possible reductions in some allergy and asthma symptoms.
So: a device that reliably does the physical job, with a health payoff smaller and less certain than the marketing implies. Preliminary is not the same as disproven. It is also not the same as proven.
Which raises the question a filter cannot answer. If you are shopping for an air cleaner because something is wrong, the machine is the wrong first stop. A cough that will not clear after three weeks. Wheeze that is new. Breathlessness on stairs you used to climb fine. Headaches that ease when you leave the house and come back when you return. Those belong to a clinician, and the last one is worth naming out loud at the visit. A child whose asthma is poorly controlled needs their treatment plan reviewed by their doctor. A filter can be part of that plan. It is never a replacement for it.
What matters most here is the order you do things in, not the product you buy. Find the source and remove it if you can. Open the house to outdoor air when the outdoor air is good. Then handle what is left. Buy a portable unit rated for a bigger room than you have and run it continuously on a quiet speed where you sleep or sit. Put at least MERV 13 in the furnace if your system can take it. Change both on a schedule you wrote down. That sequence costs less than most of what gets sold as a solution, and it is the part that moves air through a filter.
Sources
- U.S. Environmental Protection Agency. “Air Cleaners and Air Filters in the Home.” epa.gov
- U.S. Environmental Protection Agency. “Guide to Air Cleaners in the Home.” epa.gov
- U.S. Environmental Protection Agency. “Residential Air Cleaners: A Technical Summary, 3rd Edition.” epa.gov
- U.S. Environmental Protection Agency. “Ozone Generators that are Sold as Air Cleaners.” epa.gov
- U.S. Environmental Protection Agency. “Improving Indoor Air Quality.” epa.gov
- U.S. Environmental Protection Agency. “What is a MERV rating?.” epa.gov
- U.S. Environmental Protection Agency. “What is a HEPA filter?.” epa.gov
- U.S. Environmental Protection Agency. “Air Cleaners, HVAC Filters, and Respiratory Viruses.” epa.gov
- PubMed. “Riederer AM, et al. "Effectiveness of portable HEPA air cleaners on reducing indoor PM2.5 and NH3 in an agricultural cohort of children with asthma: A randomized intervention trial." Indoor Air, 2021. DOI: 10.1111/ina.12753.” doi.org
- PubMed. “Drieling RL, et al. "Randomized trial of a portable HEPA air cleaner intervention to reduce asthma morbidity among Latino children in an agricultural community." Environmental Health, 2022. DOI: 10.1186/s12940-021-00816-w.” doi.org
- PubMed. “Mudgal SK, et al. "Effectiveness of Personal-Level Protective Measures Against Air Pollution on Respiratory Health: A Systematic Review and Meta-Analysis of Randomized Controlled Trials." Indian Journal of Occupational and Environmental Medicine, 2026. DOI: 10.4103/ijoem.ijoem_328_25.” doi.org