How long do air purifier filters actually last?

The change-filter light is not measuring your filter. It is counting hours.

Every purifier page answers this question with a number of months. The machine in your room answers it with a number of hours, and those two answers diverge the moment your daily runtime is not the one the manufacturer assumed. Below is what the indicator actually measures, how far apart the two answers can get, and the filter in the box that no indicator covers at all.

What we read

We built this from published manufacturer documentation and reset procedures across the major US purifier brands, the filter-change thresholds those units use, the published replacement-interval guidance from the appliance and environmental bodies, and published CADR figures.

We did not measure a filter, and we have not run a purifier long enough to load one. Where a figure comes from a manufacturer’s documented threshold, a published interval or a certified CADR rating, we say which. One source used here has a commercial interest in filter sales and states that its pages are AI-assisted; that is labelled in the sources section rather than glossed over.

The light is a timer, not a sensor

This is the single fact that reframes the whole question, and it is not stated on most product pages.

A purifier’s filter indicator is tied to a cumulative runtime counter. A true air-quality sensor measures particulate concentration in real time. A filter timer counts how many hours the fan has run and raises the alert when that number hits a factory-set limit. Replacing the filter does not change the counter — you have to reset it manually, which is why a brand-new filter still shows a red light on a unit nobody reset.

The thresholds differ by brand, and the published figures are specific:

Note the Coway row, because it shows the two systems coexisting in one market. Some manufacturers count hours; some count calendar time; and the two are not equivalent once runtime varies.

The consequence is the sentence worth remembering: none of these timers know whether your filter is actually dirty. A household with two large dogs and heavy cooking genuinely needs a new filter sooner than the counter suggests. A single-occupant apartment with few pollution sources can run the same filter well past the alert. The light is a prompt to inspect, not a measurement.

Why “months” and “hours” cannot agree

Here is where the two answers separate, and it is arithmetic rather than opinion. Take a 2,000-hour threshold — the common one — and vary only how long the machine runs each day. This is our calculation from the published threshold:

4 h/day 500 d

8 h/day 250 d

12 h/day 167 d

24 h/day 83 d

0125250375500 days

The 2,000-hour threshold is the manufacturer’s; the calendar figures are our division of it by each daily runtime, the same arithmetic the table above sets out. The rounding is ours as well, so read the month conversions as approximate. The source figures above were read on 27 September 2026.

That is a sixfold spread in how often the same machine asks for the same filter, decided entirely by how you use it. And it explains a contradiction that shows up in owner discussions constantly: one person reports that the light came on “way too early” and another that it never came on at all. Both are describing the same timer under different duty cycles.

It also means the calendar guidance and the indicator are answering different questions. Six months is a calendar rule based on typical use. 2,000 hours is a runtime rule that ignores the calendar. If you run your purifier around the clock — which is exactly what you do if someone in the house has allergies — the runtime rule fires first, and it fires long before six months.

What the published intervals actually say

The calendar guidance most pages quote comes from the appliance industry and the environmental regulator, and it is more conditional than the headline suggests.

Two rows deserve attention. The 3–4 month figure for pet, smoke and smoker households is half the headline interval and applies to a very large share of actual buyers. And the pre-filter instruction carries a warning worth repeating: a pre-filter put back wet can grow mould inside the unit, which means the machine then blows mould into the room it is supposed to be cleaning.

The filter with no indicator at all

Now the gap that no light covers. Most purifiers use two filters: a HEPA layer for particles, and an activated carbon layer for odours and volatile organic compounds. The runtime counter is calibrated for the HEPA layer.

Carbon does not work the same way. It does not clog — it saturates. The pores fill with the molecules they have adsorbed, and once full the layer stops removing anything. There is no visible change, no airflow reduction, and in most consumer units no sensor that can detect it.

So the carbon stage typically needs replacing every 3–6 months on a schedule that has nothing to do with how clean the pre-filter looks, while the indicator light is counting toward a HEPA threshold that may be set far later. If your reason for owning a purifier is cooking smells, pet odour or VOCs rather than dust, the indicator is measuring the wrong thing entirely.

What a clogged filter actually costs you

Ignoring the indicator indefinitely is not a neutral choice, because a loaded HEPA filter resists airflow.

The published relationship is direct: a clogged filter reduces airflow, which forces the motor to work harder and lowers the effective CADR — the certified clean air delivery rate that determines how much air the unit actually cleans per minute. The figure given is that a 20% reduction in airflow can drop a 250 CADR unit to an effective output below 200 CADR.

That matters because CADR is the number that decides whether the purifier covers your room. A machine that no longer delivers its rated figure is a machine that no longer covers the square footage it was bought for — it is still running, still using electricity, and quietly underperforming. This is the honest cost of treating the light as a suggestion.

The replacement-filter problem worth knowing about

One more practical trap, and it is specific to this category. Some newer smart models use embedded chips or barcodes on the OEM filter. A third-party replacement without the chip can leave the indicator permanently lit, because the machine never registers that a filter is present.

That converts a routine consumable into a locked-in one: the cheaper compatible filter costs less per unit but may leave you with a warning light you cannot clear. It is checkable before you buy a machine — look for whether the model requires a genuine OEM filter to reset — and it is much easier to check then than afterwards.

What to check before buying

• Is the indicator a timer or a sensor? Assume timer unless the documentation says otherwise, and find out what threshold it uses — hours or calendar months.
• Do the arithmetic on your own runtime. Take the threshold in hours and divide by how long you expect to run it daily. That is your real interval, and it is probably not the one on the box.
• What does the replacement filter cost per year? Combine the interval you just calculated with the part price. In this category the annual consumable cost varies by an order of magnitude between models — see the filter cost comparison.
• Does the unit have a separate carbon stage, and is it separately replaceable? If the carbon is bonded to the HEPA in one cartridge, you replace both on the HEPA schedule and pay for carbon you may not have used.
• Does it require a genuine OEM filter to clear the light? A chip-locked model removes the option of a cheaper compatible filter later.
• Can you inspect the filter yourself? Since the light is a timer, the only real measurement available to you is looking at the filter. A unit that is easy to open and inspect is worth more than one with a more sophisticated-looking indicator.

The honest answer

A HEPA filter in a typical home lasts about six months to a year, or three to four months if there are pets, smokers or wildfire smoke. A carbon stage lasts three to six months and nothing in the machine will tell you when.

But the number your purifier will act on is a count of hours, and the calendar date it produces depends entirely on your daily runtime — anywhere from under three months to over sixteen. The light is a prompt to look at the filter, not a verdict on it, and the machine has no idea what is actually in your air.

Questions this page answers

How long do air purifier filters actually last?

It depends on runtime, not on the calendar. The common threshold is a 2,000-hour counter, which is 83 days running continuously or about eight months at eight hours a day — the same filter and the same counter, very different calendar answers.

Is the filter-change light a sensor?

No. It is a runtime counter, not a sensor. It counts cumulative fan hours and triggers at a factory threshold, typically somewhere between 2,000 and 6,000 hours.

What do the published replacement intervals say?

AHAM and EPA guidance is 6–12 months for typical homes and 3–4 months with pets, smokers or wildfire smoke. That is a calendar rule being compared against a runtime rule, which is why the two cannot agree.

Which filter does nothing warn you about?

Carbon. It saturates with odours and VOCs, the saturation is invisible, and most consumer units have no way to detect it at all.

Source documents

Each figure above comes from one of these four, read on 1 October 2026. The thresholds, the timer behaviour and the CADR relationships come from one published 2026 guide that states it analysed reset procedures across more than 40 US-market models. Two disclosures about that source: it is an Amazon affiliate site, and it states on the page that it was researched and written with AI assistance. We have used its reported thresholds and its cited sources, not its recommendations. The 6–12 month and 3–4 month intervals are attributed by that source to AHAM and the EPA, and we have kept the attribution rather than presenting the figures as the publisher’s own. The CADR definition comes from AHAM’s Verifide programme. Any figure that is our arithmetic is labelled as such.

• That the filter indicator is a runtime counter rather than a sensor; the documented thresholds per brand; that the timers do not know whether a filter is dirty; the airflow and CADR relationship; the replacement intervals; the carbon-saturation point; and the OEM chip behaviour: a published 2026 guide that states it analysed reset procedures across more than 40 US-market purifier models against manufacturer documentation and owner manuals. Two disclosures about that source: it is an Amazon affiliate site with commercial links to filter sales, and it states on the page that it is “researched and written with AI assistance”. We have used its reported thresholds and its cited industry sources; we have not used its product recommendations.
• The 6–12 month baseline interval, and the 3–4 month interval for pets, smokers and wildfire smoke: attributed by the source above to AHAM (the Association of Home Appliance Manufacturers) and the EPA. We have kept that attribution rather than presenting the figures as the publisher’s own.
• The CADR definition and the certification framework: AHAM’s Verifide programme, cited by the source above as the authority for verified CADR ratings.
• The 20% airflow reduction reducing a 250 CADR unit below 200 effective CADR: stated in the source above as a worked example. It is an illustration of the relationship rather than a measured result for a specific model, and we have described it that way.

One note on the duty-cycle table above: the calendar figures are our arithmetic from the published 2,000-hour threshold, assuming constant daily runtime and no seasonal variation. They are not figures any source states. Real use is not constant, so treat the table as showing the size of the spread rather than predicting your own date.