Allergen particle sizes, explained

This is the concept that makes the rest of the category legible. It takes five minutes and it inoculates you against most air purifier marketing.

A micron is a millionth of a meter. A human hair is somewhere around 50 to 70 microns across, which is a useful anchor: everything on this page is smaller than a hair, most of it far smaller, and none of it is visible individually.

Filtration efficiency is not a straight line where bigger is easier and smaller is harder. It has a minimum, at roughly 0.3 microns, and both larger and smaller particles are captured more efficiently than that.

Above the minimum, particles have enough mass and size that they cannot follow the airstream around a fiber — they are intercepted or impact directly.

Below the minimum, particles are small enough to be pushed around by collisions with air molecules. That random walk — Brownian motion — means they wander off the streamline and hit fibers far more often than their straight-line path would suggest.

At about 0.3 microns, a particle is too small for efficient interception and too large for much diffusion. That is the worst case, and it is why the standard tests there. So “captures particles as small as 0.3 microns” describes the hardest thing the filter does, not the limit of what it does — and marketing boasting about capturing 0.003 microns is describing an easier regime while implying a harder one.

Almost everything in the table above is comfortably within HEPA-grade media’s competence, and much of it is within the competence of a filter rated well below HEPA. That is why filter grade turns out to be a weak differentiator between decent machines, and why CADR — which combines filtration efficiency with how much air actually passes through — is the number worth comparing.

The one place particle size does change a decision is PM2.5. Both registers now publish a PM2.5 CADR computed from the smoke and dust results, and if your concern is cooking, traffic or wildfire smoke rather than pollen, that is the column to read.

There is no CADR test for pet dander, mold spores or dust mite allergen specifically. The standard uses tobacco smoke, dust and pollen because they span a useful size range and are reproducible in a laboratory — not because they are what is bothering you. CADR is a well-designed proxy and it is a proxy, which is a thing we would rather say than let you assume otherwise.