How far do bed bugs actually travel between apartments?

Shared walls and how pests move between units · reviewed August 2026

What happens when you mark bed bugs and watch where they go?

They leave, in every direction, and rather more of them do it than a building would like.

The question of whether bed bugs move between apartments is usually argued from complaint patterns, which is weak evidence because complaint patterns also reflect who complains. It has, however, been tested directly. Researchers working in an occupied low-income apartment building marked bed bugs in a set of apartments, released them, and then recovered marked individuals in the surrounding units over the following weeks, publishing the results in PLOS ONE.

The findings are worth setting out precisely, because the detail is what makes them useful.

Over an initial period of fourteen to fifteen days, fourteen marked bed bugs — seven nymphs, six females and one male — were recovered in neighboring apartments. Dispersal rates over that window ran from zero to five percent. Over a longer window of twenty-eight to thirty-two days, twelve marked bed bugs were recovered across ten neighboring apartments. Shortest-path calculations between release and recapture points put the distance covered by large nymphs and adult females at at least twelve meters.

And the direction of travel is the part that should change how buildings are inspected: marked bed bugs dispersed from the release apartment to apartments in all directions — above, below, adjacent, and across the hallway.

One more figure gives all of the above its weight. The estimated number of bed bugs per apartment in the study units ran from about 2,433 to 14,291 at four to seven days after release. A dispersal rate of a few percent sounds small. A few percent of ten thousand insects, repeated month after month, is not small at all, and it is arriving somewhere.

Why does it matter that most of the travelers were female?

Because a mated female is a founding event and a male is a passenger, and the recapture data is strikingly lopsided.

Ninety-two percent of the marked adults recovered in neighboring apartments — eleven of twelve — were females. Whatever the mechanism behind that bias, its consequence for a building is direct. The insects arriving in an apartment that has never reported anything are disproportionately the ones capable of starting a population there without any further help from the source unit.

That resolves something managing agents observe and find baffling: neighboring units do not tend to produce a trickle of odd sightings that fades away. They go quiet for a period and then produce a real infestation. If dispersal were random with respect to sex, a proportion of arrivals would be dead ends. Biased towards mated females, far fewer of them are.

It also puts a shape on the timeline. A founding female arriving in a neighboring unit is not detectable by anybody for some weeks. She is one insect in a room, her offspring are small and pale, and nobody reacts to a bite they did not notice. By the time that apartment becomes a complaint, the building has lost the window in which the ring around the original case would have been cheap to inspect.

Which is the argument for inspecting on the strength of the geometry rather than on the strength of a report, and the whole reason this cluster exists as a separate class of question from the treatment itself. The treatment side is covered on the bed bug treatment page; the question here is where to point it.

What inspection scope does that evidence actually support?

A wider one than most buildings run, and the widening is specific rather than vague.

The conventional ring is the six-neighbor cube: the unit itself, the units sharing each wall, the unit above and the unit below. It is a sensible default and it is what most managing agents have been told. The researchers, looking at their own recaptures, said so explicitly and then said it was not enough — current recommendations include the units above, below and adjacent to the known infested unit, and on the strength of their findings they recommended expanding the scope of neighboring unit inspections to include units located across the hallway.

New York City’s own guidance to owners and building managers arrives at a similar place from a different direction, telling owners that if an apartment is found to have bed bugs they should notify and inspect all units that are across, above and below the infested one.

So the working scope for a confirmed case in a corridor-plan building is the cube plus the facing units, and in practice we would add two refinements that the geometry implies.

Follow the line, not just the ring. Where units are stacked with continuous vertical elements between them, the second unit above and below is worth an interceptor even if it does not get a full inspection, because vertical routes do not stop at one floor.

Include the units nobody lives in. A vacant apartment in the ring is not a low-risk unit; it is an uninspected one with nobody to report anything, and it is frequently used for storage of exactly the furnishings that carry the problem.

The scope decision is where most of the money in a building-wide program is won or lost, and it is made before anybody treats anything. A program with a correct scope and an ordinary treatment will beat a superb treatment inside the wrong boundary every time.

Why is a corridor a route rather than a gap between routes?

Because it is the one common element every apartment on the floor opens onto every day, and because the gap beneath a dwelling door is a designed opening.

Once marked insects have been recovered across the hall, the corridor stops being neutral space. Look at what it actually offers something that travels along edges in contact with two surfaces: a continuous junction where carpet or flooring meets the wall base, running the full length of the floor and past every door; the undercut beneath each door, which exists for airflow and is not going to be eliminated; carpet edges and thresholds at each opening; and, at the ends, the lift lobby, the stair enclosure and often a refuse room.

Add the two rooms that concentrate the traffic. A shared laundry receives infested textiles from every line in the building, in bags, on the floor, on folding surfaces. A refuse room receives the furnishings people discard when they realize what they have, frequently unwrapped, often stored for days before collection. Purdue’s guidance for property managers describes the corollary at unit scale — once established, bed bugs move from room to room by crawling and using cracks and openings in walls and ceilings — and the same behavior operating in a corridor produces exactly the across-the-hall result the recapture work measured.

None of this makes the corridor the primary route. Vertical and party-wall paths remain the heavier traffic. It does mean that a scope drawn only from the shared-wall diagram is drawn from an incomplete map, and that the common areas are part of the inspection rather than the space between inspections.

What makes a population disperse in the first place?

Two things, and a building controls one of them.

The first is density, and the population estimates above are the clue. An apartment holding several thousand insects has exhausted the harborage close to the host, and individuals are pushed outward to marginal sites — behind baseboard, into the wall base, through the outlet box, under the door. Dispersal on that account is a function of how long a case has run unaddressed, which is a function of how quickly it was detected, which loops directly back to whether the building was inspecting or waiting.

The second is displacement, and it is the one buildings cause. A treatment that disperses rather than kills — a repellent product applied broadly, an aerosol used by a resident from a hardware shop, a badly executed application that hits some harborage and disturbs the rest — pushes survivors along the same routes at speed. This is the mechanism by which a building’s problem gets measurably worse in the fortnight after a resident treats their own apartment, and it is the strongest practical argument for taking the work out of residents’ hands rather than merely advising them.

There is a third factor that is not dispersal at all but produces the same map, and it should be named so it is not mistaken for the others: people move infested items. A resident who bags a mattress and carries it down the corridor to the service lift has seeded a route in one trip that the insects would have taken months to establish. That is why the preparation instructions given to residents matter as much as the treatment specification, and why the instruction not to move furnishings out of a unit is the single most important line in them.

How do you find out which units are involved without waiting for complaints?

You instrument the building, because a complaint map and an infestation map are different documents.

The gap between them is biological. A substantial proportion of people show no visible reaction to bites at all, and among those who do, the reaction can be delayed and is easily attributed to something else. So a building that treats reports as its data set is working from a sample that is filtered by human immunology and by whether a resident wants their neighbors to know.

Interceptors under the legs of beds and upholstered furniture are the practical answer at building scale, because they turn each unit into a passive monitor that works every night without anybody being asked to notice anything. Purdue’s guidance describes the mechanism plainly — insects moving to or from the bed or other furniture are trapped in the interceptor — and sets a maintenance cadence of checking every seven to ten days. Placed across a defined ring rather than only in a confirmed unit, they answer the question that determines the whole program: which units are sources and which are receiving.

Two operational notes from doing this in occupied buildings. Interceptors are only as good as the furniture arrangement, so a bed pushed against a wall or with bedding touching the floor has bypassed the instrument, and the placement visit has to include saying so. And the record has to be per unit and per date, because a single check tells you very little and a series tells you the direction of travel.

Where a building needs the answer faster or at a scale that visual inspection cannot cover, canine bed bug detection is the tool for surveying a ring of units in a morning rather than a week. Those inspections are billed per unit.

What clock is running while the scope is being argued about?

Two of them, and they run on different tracks.

The immediate one is the correction period. The Health Department’s guidance to owners and building managers states that the landlord has 30 days to correct the problem, and phrases the obligation as getting rid of the infestation and keeping the affected units from becoming infested again. That second half is doing real work. A unit cannot be kept from reinfestation while the units around it are neither inspected nor treated, so the wider scope is not a best-practice flourish; it is what the stated obligation requires in a building where the neighbors are connected.

The slower one is the reporting year. HPD’s own frequently asked questions on the bed bug annual report set out what a multiple dwelling has to file: the total number of dwelling units, the number where an infestation was found during the reporting period, the number where eradication measures were employed, and the number that became reinfested after eradication. The filing window runs from 1 December to 31 December, covering a reporting period that runs from 1 November to 31 October. The completed report goes to occupants either at lease commencement or renewal or by posting it in a prominent public location, on HPD’s own form rather than a version the building has drafted, and corporate owners including cooperatives and condominiums file electronically.

Look at the third and fourth of those fields and the connection to scope becomes obvious. Eradication measures employed and became reinfested after eradication are per-unit facts with dates attached, and a building can only state them honestly if somebody was recording, for each unit in the ring, when it was inspected, what was found, what was done and when it was verified. That record is a by-product of running the program properly and is impossible to reconstruct in the last week of December.

Local Law 55 of 2018 sits underneath both, requiring owners of multiple dwellings to inspect for and address indoor pest and allergen hazards and to keep records of what they did. The Local Law 55 and landlord and tenant responsibilities pages set the statutory picture out in full; what belongs here is the narrower observation that all three instruments assume a building-level record and none of them can be satisfied from a stack of individual work orders.

Who commissions what, and why does the scope stall?

Because the part that resolves the building is the part nobody’s lease makes anybody’s problem.

Treatment inside an apartment is comparatively easy to authorise. There is a complaining resident, a visible problem and an obvious beneficiary. Inspecting eight uncomplaining apartments and sealing penetrations in common elements has no complaining resident, no visible problem, and a beneficiary who is the building as an abstraction. It is a common-element decision, which means it belongs with the board or the managing agent rather than with any shareholder or tenant, and it competes for approval against work that is easier to explain at a meeting.

The failure mode that follows is predictable and expensive. The building treats the units that complain, one at a time, for eighteen months. Each treatment works locally. The complaints migrate. The invoices accumulate to a figure that would have paid for the ring inspection several times over, and by then the case history is long enough that the December report has an awkward number in the reinfestation column.

Two things break the pattern in our experience. The first is putting the scope decision in front of the board as a single item with a defined ring, a schedule and a closing criterion, rather than as an open-ended program. The second is planning for the residents who cannot prepare unaided — people with mobility limitations, people living with a great deal of stored material, people who are frightened of what a report will mean for their tenancy. A building that has no plan for those apartments ends up with permanently untreatable units that reseed the ring, and no amount of scope discipline elsewhere compensates.

Boards and agents will find the governance argument at length on the co-op and condo page, the duties version on board obligations, and the operator’s version for hotels on the hospitality page.

Does a small building fail differently from a large one?

Yes, and the difference is worth knowing because the small building usually thinks it is exempt.

A doorman building of a hundred and fifty units has a corridor plan, stacked lines, a managing agent, a maintenance staff and a budget cycle. Its failure mode is the one described above: correct treatment, insufficient scope, complaints that migrate along a line for eighteen months. The fix is organizational, and once the board approves a ring the machinery to execute it already exists.

A converted brownstone, a two-family in Astoria or Whitestone, or a six-unit walk-up in Park Slope fails differently. There is no corridor to speak of — often just a stair and a landing — so the across-the-hall geometry collapses to a distance of a few feet, and every unit is effectively adjacent to every other. Floor assemblies in converted stock are frequently original, which means joist ends bearing into party walls and continuous voids that were never intended to separate anything, because when the building was built it was one household. And there is nobody whose job this is. The owner may live in the building, the managing agent may be a relative, and the decision to inspect four apartments rather than one is being made by someone who will meet all four residents on the stairs tomorrow.

Small buildings therefore under-scope for social reasons rather than financial ones, and they are the properties where a single unaddressed case most reliably becomes the whole structure. If you own or manage one, the useful mental adjustment is that your ring is not a subset of the large building’s ring. It is the entire building, because at that scale it already was.

Across the city line the statutory picture changes and the physics does not. Nassau and Suffolk multi-family properties are not filing an HPD annual report and are not working to a city correction period, so the discipline has to come from the owner rather than from a deadline. What does not change is the geometry: a garden apartment complex with stacked plumbing and shared stair halls transmits along exactly the same routes, and a two-family house in Huntington Station with a converted upstairs unit has a party floor rather than a party wall and behaves accordingly. The multi-family page sets out what binds an owner on each side of that line.

What does a defensible record of scope look like?

A table, kept as you go, with one row per unit and dates in every column.

The columns that matter are the ones a reader will ask about later. Which unit. Why it entered the scope — reported, adjacent, above, below, across the hall, or vacant in the ring. The date it was notified and how. The date it was inspected and by what method. What was found, including nothing found, which is a result rather than an absence of one. What was done. The date monitoring went in and the dates it was checked. The date it was verified clear, and against what criterion. And whether it came back.

Two entries carry more weight than the rest. Nothing found in a ring unit is the entry that demonstrates the scope was real, and a program with no such rows has almost certainly only inspected apartments that complained. And an access failure, logged individually with the date, the notice given and the window offered, is the entry that distinguishes a building that could not get in from one that never tried — a distinction that is entirely documentary and entirely decisive.

None of this is bureaucracy for its own sake. It is the same record that answers a resident’s complaint, a board’s question at the annual meeting, and the four numbers HPD asks for in December, and it costs nothing beyond the discipline of writing things down when they happen.

Where does this leave one resident in one apartment?

Better informed than the building, usually, and with a short list of things that genuinely help.

Report it in writing rather than verbally, because the date matters and because the correction clock begins from a report the owner can be shown to have received. Do not treat it yourself with a retail aerosol; the dispersal it causes is the most reliable way to convert a contained problem into a floor-wide one. Do not move mattresses, upholstered furniture or bagged belongings out of the apartment, into a corridor, into storage or into a relative’s home. Put interceptors under the bed legs, because they cost little and they give you dated evidence rather than an argument. And ask your managing agent one specific question: which neighboring units are being inspected, and when.

That last question is the one that changes outcomes, because it moves the conversation from your apartment to the building, which is where the problem has been all along. The identification detail is on the bed bug profile in the pest library, the service behind this work is bed bug treatment, and everything else written in this cluster is indexed on the blog.

The work behind this piece is Bed Bug Treatment, and everything else in this cluster is indexed on the blog hub.