How HEPA filters actually work

What "true HEPA" actually certifies

True HEPA is a filtration standard requiring at least 99.97 percent removal efficiency at the single hardest particle size to capture, 0.3 micron. A filter that meets this bar is tested against that specific size, not against every possible particle size a room might contain. "HEPA-type" or "HEPA-like" filters, sold without meeting that 99.97 percent bar, are a separate and lower-performing category despite the similar name.

The 0.3 micron test point is not an arbitrary marketing number; it comes from where the filter’s combined capture mechanisms are weakest, which is covered below.

The name itself, High Efficiency Particulate Air, describes the standard rather than any single manufacturer’s brand, which is why the "true HEPA" qualifier matters when shopping: it signals the filter was actually tested to and met that named standard, rather than simply borrowing language that sounds similar. A filter box or listing that states a specific percentage removal figure at 0.3 micron is making a checkable, standard-backed claim; one that just says "HEPA-style filtration" without that number is not.

A related detail worth checking on any listing: whether the stated removal efficiency is described for the filter media itself or for the whole assembled machine. A true HEPA filter tested at 99.97 percent describes the media; the same machine’s overall particle removal in a room also depends on how much air actually passes through that media rather than leaking around it through gaps in the housing, which is a separate mechanical fit-and-seal question the media’s own rating does not answer.

How interception, impaction and diffusion actually catch a particle

A HEPA filter is not a sieve with holes smaller than the particles it catches; the gaps between its fibers are actually larger than 0.3 micron in most spots. Instead, three separate physical mechanisms trap particles as air winds through a dense mat of randomly arranged fibers.

Impaction handles the largest particles: they are too heavy to follow the air stream as it curves around a fiber, so they travel in a straight line and slam directly into it. Interception handles mid-size particles that are light enough to follow the curving airstream but still brush against a fiber’s surface as they pass close by and stick there. Diffusion handles the smallest particles, below about 0.1 micron, which move erratically due to collisions with air molecules (Brownian motion) rather than following any predictable path, and this random motion increases the odds that a tiny particle eventually collides with a fiber regardless of the general airflow direction.

The dense, randomly arranged fiber mat itself is typically pleated into a folded, accordion-like shape rather than left flat, which packs a much larger total fiber surface area into the same physical filter footprint. That extra surface area gives all three mechanisms more opportunities to catch a particle per pass of air through the media, and it is also what allows a compact filter cartridge to move a reasonable amount of air without excessive resistance despite the fine mesh of fibers involved.

Why 0.3 micron is the hardest particle to catch, not the smallest

Diffusion gets more effective as particles get smaller, because smaller particles move more erratically and randomly hit more fibers. Interception and impaction get more effective as particles get larger, because bigger particles are heavier and have more surface area to brush or collide with fibers. Between those two effective ranges sits a size window, centered around 0.3 micron, where a particle is too large to benefit much from diffusion’s random motion and too small to be reliably caught by interception or impaction.

This is why 0.3 micron is called the "most penetrating particle size" in filtration testing rather than the smallest particle a HEPA filter catches. A HEPA filter actually captures particles both smaller and larger than 0.3 micron at a higher rate than 99.97 percent; the standard tests at the one size where filtration performance dips to its lowest point, which is the honest worst-case number to certify against.

This also explains a detail that surprises people the first time they hear it: a true HEPA filter is generally more effective, not less, against particles smaller than 0.3 micron, because diffusion becomes stronger as size drops further. The 0.3 micron figure is a worst-case benchmark precisely because it is the one point on the size spectrum where neither diffusion nor the interception and impaction mechanisms are working at their individual best.

What a HEPA filter does not do

HEPA media captures particulate matter: dust, pollen, pet dander, mold spores, and smoke particles. It has no mechanism for removing gases, odors, or volatile organic compounds, which pass through the fiber mat unaffected. A purifier advertising odor or VOC reduction alongside HEPA filtration is relying on a separate activated carbon layer to do that job, not the HEPA media itself.

HEPA filtration also does nothing for carbon monoxide or radon, both of which are gases rather than particles; neither is addressed by any filter media, HEPA or otherwise.

This distinction between particles and gases is worth carrying into how a whole purifier is evaluated. A machine’s three CADR numbers, smoke, dust, and pollen, all describe particulate removal performance handled by the HEPA layer; a separate, unstandardized claim about odor or VOC reduction describes the carbon layer’s performance, tested by different, less consistently disclosed methods across manufacturers. Treating the two as one combined "air cleaning" score obscures which specific layer is actually doing which specific job.

A household weighing a purifier purchase specifically against a mold or musty-smell complaint should be especially careful with this distinction: mold spores, the particle, are captured by the HEPA layer, but the musty smell itself is a separate gas-phase compound that only a carbon layer addresses, and a purifier lacking carbon media can measurably reduce airborne mold spore counts while leaving the smell largely unchanged.

Why filter condition changes real-world capture, not the rating

A HEPA filter’s 99.97 percent rating describes a clean filter at the airflow rate it was tested at. As a filter loads with captured particles, the fiber mat actually becomes slightly more efficient at trapping additional particles, since the accumulated material adds more surfaces for interception and impaction, but airflow through a heavily loaded filter drops, which reduces the total CADR the machine delivers even as the per-pass capture rate holds or improves slightly.

This is the practical reason a purifier can feel like it is "working less well" after months of use even though the filter media itself has not gotten worse at capturing particles: the bottleneck shifts from capture efficiency, which stays high or improves, to total airflow, which drops as the media clogs. Replacing the filter on the interval covered in `/guides/how-often-to-change-a-filter/` restores the airflow side of that equation rather than the capture side, which was never really the problem in the first place.