Field Note · Shelter
·Do box fan air cleaners actually work?
The EPA put the box fan and furnace filter build in a laboratory and measured it. The answer is better than the skeptics expect, more conditional than the internet suggests, and it depends almost entirely on details people skip.
Shelter
The Conversation
Every smoke season, the box fan build goes around again.
A box fan, a furnace filter, some tape. It costs a fraction of a commercial purifier, and its replacement filters are cheaper too.[2] The claim attached to it ranges from "works as well as a purifier that costs ten times more" to "you are taping a filter to a fire hazard." Both versions circulate every year, and neither is quite right.
The reason the argument never resolves is that most people making it have not measured anything. The EPA has. Its Office of Research and Development ran these builds in a chamber against simulated wildfire smoke and published the numbers.
The findings are useful precisely because they are specific. The build works, some versions work far better than others, and the failure modes are ordinary rather than dramatic.
What Is Actually True
The design details are most of the performance.
EPA measured Clean Air Delivery Rate, or CADR, which is how much filtered air a unit actually delivers. Higher is better, and the spread between builds is wide.[1]
Box fan plus one 1-inch MERV 13 filter
The plain build, no shroud.
Same build with a cardboard shroud
The shroud alone accounts for the jump.
4-inch MERV 13 filter with a shroud
Thicker filter, more surface area.
Two 1-inch filters with a shroud
Two filters beat one thick one, slightly.
Corsi-Rosenthal cube, four filters and a shroud
The high end of what a household build reaches.
Clean Air Delivery Rate, EPA laboratory testing against simulated wildfire smoke.[1]
The shroud is the detail worth pausing on. It is a piece of cardboard, it costs nothing, and in EPA's testing it added roughly 40% to the delivered clean air.[1] Most builds posted online leave it off.
Two things decide whether any of this holds up in a real house. The first is the filter rating: EPA specifies MERV 13, because the particles in wildfire smoke are very small and lower-rated filters pass them.[1] The second is filter condition. EPA found that units running dirty filters loaded with smoke or dust were almost completely ineffective.[2] A DIY cleaner you built last August and never touched is not cleaning your air this September.
The part the internet leaves out
EPA's own position is narrower than the enthusiasm around these builds. The agency recommends DIY air cleaners "only when products of known performance, such as commercially available portable air cleaners, are not available or affordable."[3] The build is a good answer to a budget problem. It is not a free upgrade over a tested unit.
The safety question has an answer too, and it is more specific than either camp usually allows. Underwriters Laboratories tested these assemblies and found component temperatures stayed below recognized safety limits, with no observable fire hazard in the scenarios tested.[1]
That result comes attached to conditions. EPA specifies a box fan from model year 2012 or later carrying a UL or ETL safety mark, because fans built before 2012 carry known fire risks.[1]
The rest of the envelope is short. Older fans should not run unattended or while anyone is sleeping, no extension cord, and working smoke alarms in the house.[1]
What To Do Right Now
Five things, in this order.
Check the fan before you buy the filter.
Model year 2012 or later, with a UL or ETL mark on the label. If the fan in your garage predates that, it is not the fan for this.
Buy MERV 13, not whatever is cheapest.
MERV 13 is the specification, in a filter sized to your fan. EPA tested 20 by 20 inch filters, which suits the standard 20 inch box fan. Lower ratings pass the small particles that make up smoke, which is the whole reason you are building this.
Add the cardboard shroud.
The shroud is a cardboard collar taped around the fan outlet so air cannot circle back around the filter edges. It is the highest-value five minutes in the build, it is free, and it was worth roughly 40% more clean air in EPA's measurements.
Pick one room and close the house up.
EPA's guidance during a smoke event is windows and doors closed, the HVAC system set to recirculate, and one clean room where the filtration is concentrated.[3] A single unit in a closed bedroom does more than the same unit in an open floor plan.
Look at the filter, not the calendar.
EPA says change it when it looks dark brown or grey, or smells like smoke. During a heavy smoke week that can be days, not months.
Worth being clear about the limit: a unit like this cleans the room it is in. It does not make the rest of the house comfortable, and it does nothing about the air outside. One room done properly beats four rooms done partway.
One thing to skip: EPA advises using evaporative coolers and single-hose portable air conditioners sparingly in smoky conditions, because both can pull more smoke indoors.[3]
Go Deeper
Where this sits in the rest of the plan.
Indoor air is a shelter problem before it is a gear problem. The filtration only matters once the house is closed up and you have decided which room you are protecting, and the same clean-room thinking applies to an extended outage as to a smoke week. Check current conditions at AirNow, which publishes the Air Quality Index and a fire and smoke map.[4]
Go deeper
Full guidance on NWS:
Sources
- [1] U.S. Environmental Protection Agency. "Research on DIY Air Cleaners to Reduce Wildfire Smoke Indoors." Accessed September 2, 2026. [source]
- [2] U.S. Environmental Protection Agency. "Do-It-Yourself Air Cleaners: Making Cleaner Air More Accessible." EPA Science Matters. Accessed September 2, 2026. [source]
- [3] U.S. Environmental Protection Agency. "Wildfires and Indoor Air Quality (IAQ)." Accessed September 2, 2026. [source]
- [4] AirNow (U.S. EPA and partner agencies). "Air Quality Index and Fire and Smoke Map." Accessed September 2, 2026. [source]