Where wasp queens spend the winter, and why it decides next summer's nests

Stinging insects and the nesting season · reviewed August 2026

Why does one house get a nest every year and the one next door never does?

Because the two houses are not drawing from the same pool of queens, and the size of that pool is set in November rather than in May.

This is the observation that starts most of these conversations. Two properties on the same street, the same age, the same trees, the same bins, the same everything a homeowner can think of — and one of them has a colony in the soffit most summers while the other has had one nest in twenty years. People reach for luck, or for the neighbor’s fruit tree, or for something about the paint. The explanation is duller and considerably more actionable.

A wasp colony here is an annual project. It is founded in spring by a single fertilised queen who spent the winter alone in a sheltered place, and it produces, at the end of the season, a crop of new queens who disperse to do the same thing. Everything that happens in between — the workers, the nest, the traffic at the entrance, the August problem on the deck — depends on that one insect having survived from October to April, and on her having found somewhere to found a colony when she woke up.

Both of those are properties of the building. A structure with open soffit returns, an unscreened gable louver, a lifted ridge cap and a run of siding you can slide a finger behind is offering two distinct services to the local wasp population: overwintering shelter that is drier, more stable and better insulated than anything available in the woods, and, six months later, a set of ready-made cavities to nest in. Do that for a few consecutive years and the property is not merely receiving the neighborhood’s queens, it is producing them.

The house next door with a tight envelope offers neither. Its queens have to overwinter in leaf litter and under bark, where mortality is far higher, and the ones that survive have to nest somewhere else because there is no cavity available with a small defensible entrance.

That is the whole argument of this article, and it has a schedule attached to it. The number of nests a property gets next summer is decided by work done between the first hard frost and the middle of spring, and that window is currently used by almost nobody.

What actually happens to a colony in October?

It ends, completely, and only a handful of individuals carry anything forward.

Run the year out. Through the summer the colony grows on a curve that is nearly flat for weeks and then very steep — Ohio State’s extension material records nests reaching roughly four to five thousand workers and ten to fifteen thousand cells by August and late September. Penn State places the reproductive switch in the same weeks: males produced in August and September, closely followed by a brood of new queens.

Then it stops. Those new queens leave the nest, mate, and go looking for somewhere to spend the winter. The workers, the drones and the founding queen all die at the onset of cold. The nest structure itself, which took a season and an enormous amount of labor to build, is finished — Ohio State notes that abandoned nests typically decompose and disintegrate over the winter, and are only occasionally reused.

So the population crashes to a small number of mated females, distributed across the landscape, each of which is now the entire genetic and practical future of a colony.

Two consequences follow from that, and they are the reason this piece exists.

The first is that removing a nest in October is close to pointless as prevention. The colony inside it is already dead or dying, and its queens left before you got there. There are good reasons to take a nest down — a bald-faced hornet nest hanging over a walkway is unsightly and eventually messy — but nobody should imagine it reduces next year’s pressure. The insects that decide next year are already in the wall.

The second is the useful one. Between the collapse of the colony and the emergence of the queens in spring, there is a period of several months during which the property contains next year’s entire wasp population, stationary, dormant, and concentrated in a small number of predictable places. Nothing else in this trade offers that.

Where do the queens actually go?

Into anything dry, sheltered and stable — and a building is the best version of all three within a mile.

The extension literature gives the natural sites first, and they are consistent across states. Ohio State lists hollow logs, under the loose bark of dead trees, soil cavities and other protected places. Penn State names under the bark of dead trees or logs, and leaf or forest litter. Those are the classic overwintering habitats and most queens use them.

But the same sources keep naming structures alongside them, and it is the structural list that matters to a homeowner. The University of Maryland’s material on social wasps includes wall voids in its list of overwintering sites, next to leaf litter, tree bark, compost piles, logs and soil cavities. NC State’s publication on paper wasps in and around structures is the most explicit of the lot: overwintering wasp queens seek shelter in hollow trees, under bark, in wood piles, attics, chimneys, barns and under siding. Ohio State, discussing yellowjackets specifically, records that queens may sometimes be found overwintering in homes.

Line those up against the details on an ordinary frame house on Long Island and the overlap is nearly total.

Attics, which are the premium option: large, dry, dark, undisturbed for six months at a stretch, and warmed from below by the building.

Wall voids, entered at any gap in the cladding, and thermally buffered on one side by a heated room.

Behind siding and cladding, which is the most abundant habitat of all. Vinyl and aluminum siding hangs on the wall with a continuous ventilated space behind it, entered at every J-channel, corner post, utility penetration and butt joint.

Soffit and eave cavities, reached at the return, the fascia joint or a torn vent screen.

Chimney chases, particularly the framed hollow columns on newer construction, which are large empty shafts with a rusting pan on top.

Shutters, trim boards, porch columns, window casings and the gap behind a mounted light or a house number. Each of these takes one queen, and one queen is all a nest requires.

The reason a building beats leaf litter is not comfort, it is stability. Look at what the natural sites in those extension lists have in common — under bark, inside a hollow log, down a soil cavity — and it is that each of them is dry and sheltered from the weather. A leaf-litter site is neither for long: it alternates between frozen and thawed through a New York winter and takes on water whenever it rains. An attic or a wall void does neither of those things, and it sits against a heated building besides. So a property with a well-sealed envelope leaves its queens with the shelter the woods provide, while a property with an open one is offering something better than the woods, several hundred times over.

Why do wasps appear inside the house in autumn and again in February?

Because a queen sheltering in the wall navigates by light, and from inside a wall the light is coming from the wrong side.

This is one of the most common calls in this trade and one of the least well explained. A household with no wasp problem at all, in November, finds two or three large wasps crawling slowly on a windowsill or a carpet. Then nothing for six weeks. Then a mild day in February and there is another one, moving badly, walking rather than flying, apparently from nowhere.

NC State’s material sets out the mechanism plainly. A queen resting in an attic, wall void or crawlspace becomes attracted to light coming through a gap at a baseboard, around a wall fixture or around an air conditioning vent, and emerges inside the building. Once inside a dwelling the wasps are found crawling on the floor or furniture, or at windows where light is showing through. And on any warm day, the publication notes, the wasps may become active and fly about.

Read those two sentences together and the pattern in the house makes sense. In autumn, queens are entering the structure in numbers and some of them get the geometry wrong, ending up on the room side of a partition instead of in the void. Through the winter they are dormant. And then, on a warm spell, the ones sheltering in a cavity that is warmed by the building — the attic over a heated ceiling, the wall void behind a radiator, the void behind south-facing siding on a bright afternoon — reach activation temperature weeks before anything outdoors does. Some of them then head for the light and arrive in the living room in February.

The wasps behave oddly because they are not doing anything wasp-like. They are cold, they have not fed since autumn, they have no nest and no colony, and they are not defending anything. They are also, individually, of no consequence. What they are is a symptom with a specific meaning, and the meaning is worth reading: a wasp on the inside of a February window is direct evidence that the cavity it came from is open to the outside, and that the building was used as an overwintering site in the autumn.

That single observation is a better predictor of next summer’s nest count on the property than anything you can see from the driveway in July.

Where does a founding queen go when she wakes up?

Looking for a hole somebody else made, which is why the ground-nest problem recurs on the same lawns year after year.

Ohio State’s material describes the sequence: queens emerge during the warm days of April and early May, feed on nectar and insects, and then select a nest site — often an abandoned mouse or rabbit burrow, though other secluded cavities work. Penn State’s list for the eastern yellowjacket names forest floors, rock walls, creek banks and lawns, and notes that in northern urban areas nesting occurs within the wall voids of structures, including homes.

Two things are worth pulling out of that.

The first is that a founding queen is not an excavator. She has several weeks of solitary work ahead of her, raising the first brood alone with no workers to help, and no capacity at all to dig a colony-sized chamber out of soil. So she takes over an existing void. Everything she nests in was made by something else — a rodent, a frost heave, a settled paver, a rotted stump, a carpenter, or the gap left where a deck post was cut off at grade.

The second follows directly, and it is the part that belongs on a building-focused site. The supply of ground-nest sites on a property is a legacy of the property’s rodent history. A lawn with a burrow system under it — active or abandoned, Norway rat, chipmunk, vole or woodchuck — is holding a stock of pre-built yellowjacket chambers with a small defensible entrance and a dry cavity behind it, which is the exact specification. That is why some properties produce a ground nest under the same corner of the lawn most summers, and why others never produce one at all. It is not the grass. It is what is under the grass, and it is a question about rodent pressure as much as about wasps.

The rest of the inventory is hardscape and clutter, and it reads like a survey of things nobody looks under. The void beneath a patio slab or a set of concrete steps where the fill has settled. The hollow behind landscape timbers and retaining sleepers. Under a shed floor. Inside a compost bin. The space under a deck where the skirt has come away. Inside a disused grill, a rolled tarp, an upturned wheelbarrow or a stack of empty pots. A queen surveying in early May will test all of them.

The durable version of the fix is therefore the same argument in a different medium. Correcting the burrow systems and closing the voids under hardscape removes the sites themselves, permanently, in a way that treating a nest each August does not. It is also work best done outside the season, for the obvious reason: probing around a settled slab in July is a good way to find a colony with your hands.

There is one more consequence of the fact that queens do not dig. A nest found in June is in a cavity that existed in April. If you find one, look at what made the cavity, because the answer is usually still there and usually still producing more of them.

Why is a nest in May trivial and the same nest in September a serious job?

Because the colony curve is exponential and the calendar is doing all of the work.

Penn State places nest development by overwintered queens in May or early June depending on spring temperatures, with the first brood of workers emerging in June, after which the founding queen stays inside the nest permanently. At that point the colony consists of a single insect and a comb the size of a coin, and the entire colony can be ended by knocking it down with a broom handle. There is no defensive force, because the workers do not exist yet.

Fourteen weeks later the same colony, in the same soffit, is the four to five thousand workers Ohio State describes, occupying a structure a couple of feet across, and defending it. Nothing about the site changed. Nothing about the species changed. The date changed.

The practical asymmetry is worth stating in operational terms rather than as an aphorism. In May the job is done from a step ladder in daylight with no protective equipment and takes ten minutes. In September the same job is done after dark, in a veil, from a position chosen because it is not directly under the entrance, with the household kept indoors, and it carries a real risk of multiple stings if anything is misjudged. The cost, the risk and the disruption all scale with the population, and the population scales with how long nobody looked.

There is a further point that follows for a wall-void colony specifically, and it is why the calendar and the envelope belong in the same discussion. A small colony in a stud bay in June is contained by the cavity it is in. A very large one in September has expanded to fill the bay and is pressed against the back of the interior finish, which is the situation in which heat, vibration or a badly judged intervention brings workers through a ceiling fixture or a switch box into the room. The difference between those two situations is three months of nobody walking the eaves.

Twenty minutes with a pair of binoculars in the second half of May is the highest-value unpaid work available on this subject, and it is worth more than anything anyone can sell you in August.

What is the durable fix, and when is it done?

Closing the cavities, and doing it between the first sustained cold and the middle of spring.

That timing is the argument. Almost all wasp work in this market happens in August and September, because that is when the phone rings, and by then every option is a bad one: the colony is at maximum size, the site is hazardous, and anything sealed has to be sealed around a treatment. Whereas in December the colonies are dead, the queens in the structure are dormant, the ladder work is safe, and closing an opening does not risk pushing anything into the house. There is no active entrance to worry about, because there is no active colony.

The inventory is the same one that governs everything above the gutter line, which is not a coincidence — a nesting queen and a nesting squirrel are shopping for the same specification.

Soffit returns. The boxed triangle where the roof edge dies into the wall at a gable end, which is open into the eave or attic on an enormous proportion of houses here and is the most productive wasp cavity we find.

Gable louvers. Re-screened from behind with metal, never blocked, because they are there to move air.

Ridge vents. End caps reseated, torn mesh replaced, compressed baffles dealt with.

The soffit-to-fascia joint and fascia gaps, particularly where vinyl soffit was fitted over an older wood assembly and is held in a channel rather than fastened along its length.

Siding terminations. J-channel, corner posts, the top course under the frieze, and every utility penetration through cladding.

Weep holes in brick veneer, which must keep draining and are therefore screened rather than filled.

Attic and gable vents generally. NC State’s publication makes the same recommendation from the other direction: in homes, attic vents should be properly screened to exclude overwintering queens.

The materials are the ones this firm uses on everything: custom-fabricated 26-gauge galvanised sheet metal, cement and mortar on masonry substrates, Xcluder fill fabric and door sweeps where an opening has to stay serviceable, copper and stainless mesh and hardware cloth where air has to keep moving, and mechanical fasteners throughout. Expanding foam does not appear in any role. It is the trade’s default closure and its most frequent failure, and once an opening has been foamed it looks finished in a photograph and stops being inspected.

What this buys is not a wasp-free property, and nobody should claim one. Wasps will still nest in the ground, in the shrubs, under the deck and on the neighbor’s shed. What it removes is the building’s contribution: the overwintering shelter that raises local queen survival, and the cavities with small defensible entrances that turn a passing queen into a colony inside your wall.

What should you actually do, and in what order?

Work backwards from the season you are in, because the right action in October is not the right action in June.

Late summer and early autumn, with an active nest. Deal with the colony first and leave the opening alone in the meantime. Sealing an entrance while there is traffic at it is the classic self-inflicted injury on this subject, and it is covered in full on the wasp and hornet removal page. Note where the entrance is and how heavy the traffic is; both are useful and neither requires going near it.

Autumn, after the colony has finished. This is when the survey should happen. Walk the elevations with binoculars and write down every gap: soffit returns, louvers, ridge ends, fascia joints, siding terminations, shutters, trim. If wasps have been appearing inside the house, note which rooms — that narrows the cavity considerably.

Winter. Do the work. This is the window, and it is empty because nobody thinks about wasps in January. The closure done now removes the queens sheltering in those cavities and removes the nest sites in the same operation.

Mid-spring. Watch for founding queens surveying the roofline, which is a visible behavior: a single large wasp working slowly along an eave, entering and leaving gaps. Anything she gets into is a gap that was missed.

Late May and June. Twenty minutes with binoculars every couple of weeks. Any starter comb found now is a coin-sized problem.

One thing worth repeating, because it changes the answer completely: if what you have is honey bees rather than wasps, none of the above applies and the correct call is a beekeeper. Identification detail sits in the pest library on the eastern yellowjacket, bald-faced hornet and paper wasp profiles.

The rest of the writing on the stinging-insect year is indexed on the blog, the sealing trade underneath all of it is structural exclusion, and the same roofline details that host a colony in August host wildlife in March.

If your house produces a nest most years, the useful conversation is not about this year’s nest. Tell us where they went in and what month you last saw wasps indoors — the consultation is free, and winter is the season this problem is actually solved in.

The work behind this piece is Wasp & Hornet Removal, and everything else in this cluster is indexed on the blog hub.