Air Leakage Around Exterior Doors: What Actually Works
The most common mistake I see on residential weatherization calls is treating door perimeters like window openings. They're not. Doors move. They swell with humidity. The frame shifts when the building settles or when someone actually closes the door, and that movement is what breaks traditional foam-and-caulk setups over time. Door Guided Air Sealing is the approach that accounts for that reality by creating a seal that accommodates the door's operational range rather than fighting it. The core idea is straightforward but easy to mess up if you don't think through the assembly order. You're building an airtight transition between the rough opening and the door jamb, then separately sealing the gap between the finished jamb and the door itself. Most people lump these into one continuous gap and fill it with expanding foam, which works until the door warps or the foam degrades. The guided seal separates the structural plane from the moving plane. Here's the sequence. First, remove the interior trim and the casing entirely so you can see the full depth of the rough opening to jamb gap. That gap varies by manufacturer, but it's typically between 3/8 inch and 5/8 inch on modern prehung units. Stuff rigid insulation or fiberglass backer rod into that space from the interior side. It should be tight enough that it doesn't fall out when you push it, but not so compressed that it bows the jamb inward. This creates your primary air barrier at the rough opening plane.
Then you seal the gap between the jamb and the finished door using a combination of perimeter gaskets and low-expansion foam applied only to the hinge and latch sides where movement is minimal. The meeting stile of double doors and the bottom threshold area get special attention because those are where most leakage actually occurs. I use a T-adhesive foam sealant rather than liquid-applied sealants on the interior perimeter. Liquid sealants crack within two seasons on exterior doors in climates with real temperature swings.
Why People Get This Wrong
I worked a house last spring where the previous installer had used fast-expanding great stuff around the entire perimeter of a steel entry door. The foam had pushed the hinge side of the jamb inward by about an eighth of an inch. The door wouldn't latch unless you lifted it slightly. The air leakage was roughly 2.3 CFM at 50 Pascals based on a blower door test, and the homeowner was getting cold drafts on the entry side even in summer. We cut out the foam, re-shimmed the hinges, and did a proper guided seal. The leakage dropped to 0.4 CFM. Another pattern I see constantly: people caulk the exterior casing-to-siding interface but leave the interior side unsealed. That exterior caulk acts as a vapor barrier on the wrong side of the assembly during heating season, pushing moisture into the framing cavity where it condenses on cold sheathing. Seal the interior first. Then do the exterior with a flexible urethane caulk rated for at least twenty-five years of UV exposure. Silicone fails here because paint won't adhere to it, and you'll need to repaint the casing eventually.
Get the Full Details
Door Guided Air Sealing for Existing Buildings
If you're dealing with an existing door that was installed before these standards existed, you still do the same thing, but the access problem is worse. You can't easily remove the trim without damaging it, so you work from the exterior where possible. Remove the exterior casing beads carefully with a utility knife and a flat bar. Clean out old failed sealant. Install a continuous bead of foam-backed door gasket tape against the stop on the interior side before you re-hang the door. For the threshold, a sweep with a neoprene bulb gasket is more durable than the cheap felt strips that come factory-installed on most replacement doors. I've found that using a combination of a pneumatic door bottom seal and a magnetic threshold seal on older wood doors reduces air infiltration more than any other single modification. These are available through most HVAC supply houses. A good door bottom like the Adams Rite Series 110 costs about eighteen dollars and seals a gap up to half an inch. The magnetic threshold is another fifteen dollars. Together they handle the bottom gap, which accounts for roughly forty percent of door-perimeter leakage on average.
Limitations and When This Approach Fails
Door Guided Air Sealing does not solve every problem. If the door frame itself is bowed or twisted, no amount of sealing will fix the underlying issue. You need to shim and plumb the frame first. Check the gap around all four sides with a Feeler gauge before you start sealing. If the variation exceeds 1/16 inch at any point, the frame needs adjustment. Sealing over a misaligned frame just moves the problem inside the wall. Heavy storm doors can also defeat a well-sealed primary door if the gaps aren't coordinated. The storm door creates a pressure differential that forces air through the primary door's relief gaps. If you're sealing the main door to under 0.5 CFM, make sure the storm door has its own weatherstripping and isn't just relying on a thin brush seal. I once spent three hours sealing a door perfectly only to have the blower door reading show no improvement because the storm door was leaking eight CFM at the same pressure. For high-performance builds targeting Passive House or similar standards, door Guided Air Sealing alone isn't sufficient. You need a continuous air barrier that wraps around the entire rough opening, and the door assembly has to be tested as a unit. Standard CA and R-values don't apply here. The air leakage rate for a certified door assembly is typically measured in the 0.3 CFM range at 75 Pascals, and achieving that requires factory-sealed units or a custom-built door frame with integrated gasket channels.
Materials I Actually Use
Wadding Strip: 3/4 inch thick closed-cell polyethylene, R-value around 2.5 per inch. Compresses to fill gaps between 1/2 and 3/4 inch. Easy to cut with a knife. Doesn't absorb moisture. About three dollars per linear foot at supply houses. Low-Expansion Foam: Soudal Big GapFill or equivalent. Expands about three times its liquid volume but stays soft enough to trim. One can covers roughly ten linear feet of a one-inch gap. Cost is around fourteen dollars per can. Do not use high-expansion foam near door jambs unless you want to re-do the work. Tape Sealants: ProTape or similar butyl-based flashing tape. Applied to the jambs before the door is hung. Creates a permanent seal between the jamb and the rough framing. Overlaps must be at least two inches. Tape alone won't fill gaps wider than a quarter inch, so combine it with backer rod for larger openings.
Interior caulk: A one-part acrylic latex caulk rated for movement. Brands like Gorilla or DAP work fine. Avoid the cheap stuff that cracks after one winter cycle. A caulk gun and a fingertip seal work better than a tool for most applications. Tools create ridges that trap dust and prevent proper adhesion.
Verification
The only way to know if your sealing worked is to test it. A handheld thermal camera shows cold spots around poorly sealed doors, but it's qualitative at best. A blower door test with the door cycling closed and open gives you a real number. If you don't have access to a blower door, a simple candle test on a windy day will reveal most gross leaks. Hold the flame near the sealed perimeter and watch for deflection. Smoke sticks are more reliable but cost more. What I found during a recent audit of a 1978 tract house: the original entry door had no perimeter seal, just a few nails and some deteriorated weatherstripping. The blower door showed 4.7 CFM leakage at that single door. After installing a new door with a guided seal assembly, the reading dropped to 0.6 CFM. That's a real difference in heating load, probably around 12,000 BTU per hour at design conditions in a cold climate. Over a full heating season, that's meaningful energy savings, though the payback period is measured in decades, not months. The bottom line is that Door Guided Air Sealing is the right approach for new construction and major replacements, but it requires understanding how doors move and designing the seal to accommodate that movement. Most failures come from treating doors as static objects and sealing them like walls. Once you respect the difference, the process is fairly quick. A standard exterior door takes about twenty minutes from tear-out to final seal if you've done this before. First time, plan on forty-five.
