A HEPA filter removes at least 99.97% of particles at 0.3 microns. That is the standard, and the EPA notes the acronym — high efficiency particulate air — is officially defined by the US Department of Energy.
“HEPA-type”, “HEPA-like”, “HEPA-style” and “99% HEPA” are not that standard. Those phrases exist because the filter does not meet it. The wording is the specification, and the difference it hides is bigger than the decimal point suggests.
The arithmetic of the gap
99% and 99.97% sound almost identical. Look at what passes through instead of what is caught and they are not close at all.
| Stated efficiency | Particles caught per 10,000 | Particles passing through | Relative to true HEPA |
|---|---|---|---|
| 99.97% (true HEPA) | 9,997 | 3 | 1× |
| 99% | 9,900 | 100 | 33× more |
| 95% | 9,500 | 500 | 167× more |
| 85% | 8,500 | 1,500 | 500× more |
A “99% HEPA-type” filter lets through roughly thirty-three times as many particles per pass as a true HEPA filter. Over an hour of recirculation that compounds, because each pass leaves a larger residue for the next pass to work on.
This matters most for the particles you cannot see. Large particles settle out of the air on their own; the fine fraction that stays suspended for hours is exactly where the efficiency difference bites.
Why the standard is written at 0.3 microns
The obvious assumption is that 0.3 microns is the smallest particle a HEPA filter catches, and that anything finer sails through. The opposite is true.
The EPA is explicit that 0.3 microns is the worst case — the most penetrating particle size — and that particles both larger and smaller than that are captured with even higher efficiency.
The reason is that a fibrous filter catches particles by several mechanisms at once, and they have different size preferences. Larger particles are intercepted and caught by impaction as they cannot follow the airstream around a fiber. Very small particles move erratically by Brownian motion, which makes them collide with fibers readily. Around 0.3 microns, particles are too small to be reliably intercepted and too large to diffuse much, so they thread between the fibers more successfully than anything else.
The standard is set at the hardest size on purpose. A filter rated 99.97% at 0.3 microns is doing better than that at 0.1 microns and better again at 1 micron.
How to read filter labels
| What the label says | What it means |
|---|---|
| True HEPA / HEPA / H13 / H14 | Meets or exceeds the 99.97% at 0.3 micron threshold. H13 and H14 are European classifications at or above that level. |
| HEPA-type, HEPA-like, HEPA-style | Does not meet the standard. No defined efficiency; could be anything. |
| 99% HEPA, 99.5% HEPA | Does not meet the standard. The number tells you how far short. |
| H11, H12 | European classifications below the HEPA threshold — roughly 95% and 99.5% at the most penetrating size. |
| “Captures 99.97% of particles” with no size given | Incomplete. Without the particle size the figure is unverifiable, and a large size makes it trivial. |
That last row is the most common evasion. Any competent filter removes 99.97% of 10-micron particles. The claim only means something attached to a size.
When a true HEPA filter is not the right choice
Two cases, and both come back to the fan.
In a forced-air HVAC system. A true HEPA filter has high resistance to airflow. Residential blowers are not designed to pull air through one, and fitting a HEPA filter into a return slot can reduce airflow enough to harm the system and the heating or cooling it is supposed to deliver. The EPA lists HEPA separately from the MERV scale for this reason, and recommends a filter rated at least MERV 13, or the highest rating the system’s fan and filter slot can accommodate. Whole-house filtration is a MERV question, not a HEPA one.
Where airflow has been traded away to get the rating. Since clean air delivery depends on efficiency and airflow together, a HEPA filter in a unit with a weak fan can deliver less clean air than a slightly less efficient filter in a unit with a good one. This is precisely what CADR measures, and it is why the CADR figure outranks the filter label when the two seem to disagree.
So the order of priority is: certified smoke CADR first, then confirm the filter is genuinely HEPA rather than HEPA-adjacent. A unit with a strong CADR and a true HEPA filter is the pairing you want. A HEPA label with no CADR published is a weaker signal than it appears.
Practical checks before you buy
- Read the exact words. “HEPA” and “True HEPA” are fine. Any qualifier in front of or attached to the word is a downgrade.
- Demand the particle size. An efficiency percentage without a micron figure is not a specification.
- Check the replacement filter is also HEPA. Some units ship with a true HEPA filter and sell cheaper non-HEPA replacements alongside it. Look at the part you will be buying for years.
- Check whether the pre-filter is separate. If the coarse mesh is bonded to the HEPA pleats, hair and lint will shorten the life of the whole assembly.
- Ignore “washable HEPA”. Washing a pleated HEPA medium damages the fiber structure that produces the rating. A washable pre-filter is useful; a washable HEPA filter is a contradiction.
Related guides
- How to choose an air purifier: the full guide
- CADR explained: the figure that outranks the filter label
- MERV ratings explained, for whole-house filtration
- Activated carbon filters: what HEPA cannot do
Sources: EPA, What is a HEPA filter? for the 99.97% at 0.3 micron definition, the DOE attribution and the most-penetrating-particle-size point; EPA, What is a MERV rating? for HEPA being rated outside the MERV system and the MERV 13 recommendation. The pass-through comparison table is our own arithmetic.