Rodent Season 2026: When Mice Move Indoors
Everybody tells you mice come inside in the fall. Almost nobody tells you when, and the one number everybody does quote turns out not to be published by the agency it is credited to. Here is a sealing date for every state, derived from NOAA overnight low normals, the gap sizes agencies actually publish, and which rodents your state has.
The thing everyone says, and what the agencies actually publish
“A mouse can fit through a hole the size of a dime.” It is one of the most repeated facts on the American internet and it is nearly always credited to CDC. CDC’s seal up page and its 2025 rodent brochure mention no dime, nickel, quarter or any other coin. Here is what CDC does publish, which is better, because you have one in a drawer:
“Check inside and outside for large and small gaps or holes. Mice can fit through a hole the width of a pencil (1/4 inch or 6 millimeters in diameter).”CDC, How to Seal Up to Prevent Rodents, page dated April 8, 2024 source
A quarter of an inch. Six millimetres. The width of a pencil. And CDC publishes no rat figure at all, on that page or in the brochure, so every rat coin size attributed to CDC has been made up somewhere along the way. The other published dimensions come from California IPM and the federal wildlife damage handbook.
In CDC’s own words: a mouse fits through a hole the width of a pencil, a quarter of an inch or six millimetres across. California IPM gives the same quarter inch for the height of gap a house mouse squeezes under.
A second and larger figure, for width rather than height, from California IPM. Almost everything written about mice collapses the two into one number. They are not the same measurement and the difference decides whether a repair works.
California IPM names young rats of either species. CDC publishes no rat figure at all, on the seal up page or in its brochure, so nobody should be attributing one to CDC.
The tightest number of the set, from the federal wildlife damage handbook, and it is about how the building was put together rather than about the animal.
Every state, by where its cool down stands today
Derived from NOAA normalsFrom the first 50F nights at Sioux Falls to the first freezing ones. Where the runway is that short, the sealing work belongs in August.
Their reference station’s normal overnight low never reaches 32F. That does not mean it never freezes there. It means the thirty year average night does not.
Orlando and Honolulu never see a 50F normal overnight low in any month, so Florida and Hawaii have no autumn deadline; their pages cover the species and the sealing work instead.
Where every state stands right now
Cooling nowPast the sealing deadline and inside the cool down: nights are already falling toward and through 50F
Deadline within 3 weeksSealing deadline lands within about three weeks
Not yetSealing deadline is more than three weeks away
Cool down overNights are past freezing and the autumn influx is over for this year
No cool downThe normal overnight low never reaches 50F, so there is no autumn cool down to date
Sealing dates, every state A to Z
The 50F and 32F columns are measurements: the dates a state’s reference station reaches those normal overnight lows in the NOAA 1991 to 2020 climate normals. The “seal by” column is ours, three weeks ahead of the 50F crossing, because no agency publishes a deadline and a calendar without one changes nobody’s behaviour. The roof rat column is the research-grade iNaturalist record for the state, which is a check on the published range rather than a range map of its own.
| State | Seal by | Nights hit 50F | Nights hit 32F | Roof rat records | Today |
|---|---|---|---|---|---|
| Alabama | October 4 | October 25 | Never | 27 | Cooling now |
| Alaska | July 27 | August 17 | October 11 | 0 | Cool down over |
| Arizona | November 2 | November 23 | Never | 35 | Deadline within 3 weeks |
| Arkansas | October 3 | October 24 | December 27 | 11 | Cooling now |
| California | October 8 | October 29 | Never | 1,550 | Cooling now |
| Colorado | August 20 | September 10 | October 28 | 2 | Cooling now |
| Connecticut | September 2 | September 23 | November 21 | 0 | Cooling now |
| Delaware | September 14 | October 5 | December 4 | 0 | Cooling now |
| Florida | No date | Never | Never | 676 | No cool down |
| Georgia | October 8 | October 29 | Never | 51 | Cooling now |
| Hawaii | No date | Never | Never | 103 | No cool down |
| Idaho | August 29 | September 19 | November 16 | 2 | Cooling now |
| Illinois | September 5 | September 26 | November 17 | 0 | Cooling now |
| Indiana | September 9 | September 30 | November 24 | 0 | Cooling now |
| Iowa | September 6 | September 27 | November 13 | 0 | Cooling now |
| Kansas | September 16 | October 7 | November 21 | 0 | Cooling now |
| Kentucky | September 21 | October 12 | December 7 | 0 | Cooling now |
| Louisiana | November 18 | December 9 | Never | 78 | Not yet |
| Maine | August 27 | September 17 | November 12 | 0 | Cooling now |
| Maryland | September 13 | October 4 | December 1 | 0 | Cooling now |
| Massachusetts | September 15 | October 6 | December 5 | 0 | Cooling now |
| Michigan | September 6 | September 27 | November 23 | 0 | Cooling now |
| Minnesota | August 31 | September 21 | November 5 | 0 | Cooling now |
| Mississippi | October 4 | October 25 | Never | 26 | Cooling now |
| Missouri | September 12 | October 3 | November 20 | 1 | Cooling now |
| Montana | August 19 | September 9 | November 1 | 0 | Cooling now |
| Nebraska | September 3 | September 24 | November 10 | 0 | Cooling now |
| Nevada | October 17 | November 7 | Never | 27 | Deadline within 3 weeks |
| New Hampshire | August 20 | September 10 | October 31 | 0 | Cooling now |
| New Jersey | September 19 | October 10 | December 8 | 0 | Cooling now |
| New Mexico | September 15 | October 6 | November 21 | 2 | Cooling now |
| New York | September 25 | October 16 | December 16 | 0 | Cooling now |
| North Carolina | September 22 | October 13 | December 14 | 18 | Cooling now |
| North Dakota | August 20 | September 10 | October 25 | 0 | Cooling now |
| Ohio | September 11 | October 2 | November 29 | 0 | Cooling now |
| Oklahoma | September 30 | October 21 | December 8 | 1 | Cooling now |
| Oregon | September 7 | September 28 | Never | 253 | Cooling now |
| Pennsylvania | September 20 | October 11 | December 9 | 0 | Cooling now |
| Rhode Island | September 8 | September 29 | November 29 | 0 | Cooling now |
| South Carolina | October 1 | October 22 | Never | 26 | Cooling now |
| South Dakota | August 25 | September 15 | October 27 | 0 | Cooling now |
| Tennessee | September 21 | October 12 | December 11 | 15 | Cooling now |
| Texas | October 16 | November 6 | Never | 663 | Deadline within 3 weeks |
| Utah | September 2 | September 23 | November 12 | 2 | Cooling now |
| Vermont | August 28 | September 18 | November 13 | 0 | Cooling now |
| Virginia | September 20 | October 11 | December 13 | 24 | Cooling now |
| Washington | September 5 | September 26 | Never | 204 | Cooling now |
| West Virginia | September 12 | October 3 | December 4 | 0 | Cooling now |
| Wisconsin | September 6 | September 27 | November 16 | 0 | Cooling now |
| Wyoming | August 13 | September 3 | October 23 | 0 | Cooling now |
Never in the 32F column means the thirty year average overnight low does not reach freezing at that station, not that the place never freezes. Reference stations are the largest metro of the three this site carries per state, which is why Arizona is anchored on Phoenix rather than Flagstaff and Nevada on Las Vegas rather than Reno.
What the observation record shows, month by month
The timing claim can be checked against an independent record: research-grade iNaturalist observations, which are photographs of a real animal in a real place, verified by other people. Two series per animal. Grey is the raw monthly count. Dark is the same count divided by all research-grade observations in the same month, which corrects for the fact that far more people are outdoors with a camera in June than in November.
For the house mouse, the raw count peaks in Oct and the effort adjusted index peaks in Nov, on 3,593 records.
The house mouse result is the positive one. Raw records peak in Oct, the adjusted index peaks in Nov and bottoms out in Jun at about a fifth of the peak. Both views of the data point the same way, which is what makes it worth believing. The animal that published guidance says moves indoors in autumn is also the animal whose observation record concentrates in late autumn.
The Norway rat record shows no autumn peak. Its raw record peaks in May, in spring. Only the effort adjusted version moves it to Dec, and when two views of the same data disagree that hard, the signal is not there. Rats get no autumn story on this site, which matches the published guidance, which describes autumn entry for house mice and says nothing of the kind about rats.
One confound. The effort denominator is dominated by plants and insects, which only exist in summer, so any winter active animal’s share of the record rises in winter partly for that reason alone. The adjusted curve on its own is weak evidence. It is the agreement between a rising raw curve and a rising adjusted curve that carries the house mouse result.
Research-grade iNaturalist observations in the United States, all years, snapshot taken 2026-08-28. Aggregate counts only; no observation photographs are used anywhere on this site, because most of them carry noncommercial licences an ad supported site cannot honour. Each series is scaled to its own maximum, because the two are in different units and the comparison is between shapes. Neither series is a population estimate.
One sentence of guidance, and three limits on it
One published agency statement puts rodent entry on a calendar. It names nighttime temperature and nothing else, and it is narrower than the internet version in three ways.
What it says. Where there is a house mouse population living outdoors nearby, the mice frequently move into buildings in autumn as the overnight temperature falls. Nighttime temperature is the cue the fact sheet names, and it names no other. UC Statewide IPM Program, Pest Notes: House Mouse, UC ANR Publication 7483, revised July 2025. Read it there.
Limit one, the conditional. The same fact sheet attaches a condition that matters: this is about mice that are living outdoors and come in. It also lists a route into buildings that has nothing to do with weather at all, namely arriving inside packaged goods and delivery vehicles. UC Statewide IPM Program, Pest Notes: House Mouse, UC ANR Publication 7483, revised July 2025. Read it there.
Limit two, this is about mice. The companion fact sheet on rats says nothing at all about rats moving indoors in autumn, or about cold driving them in. Rat guidance is written around harborage, food and structural access instead. UC Statewide IPM Program, Pest Notes: Rats, UC ANR Publication 74106, revised June 2025. Read it there.
Limit three, cold is not the only seasonal driver. Arizona extension reports that roof rats are more commonly reported in the cooler months, and attributes it to backyard citrus rather than to the cold: tangerines can be heavy on the tree as early as September and oranges run on into spring. University of Arizona Cooperative Extension, Roof Rats: Identification, Ecology, and Signs, az1775, July 2018. Read it there.
Why overnight lows and not first frost dates
The guidance does not say frost. It says nighttime temperature. So this page reads the NOAA 1991 to 2020 normal daily minimum temperature, day by day, at a reference station in each state, and finds the date it falls through 50F, rather than reusing the first freeze normals carried elsewhere on this site.
The 50F threshold is ours. No agency publishes a number, so this site uses one that lands inside the September to November period the guidance describes and sits far enough above freezing to arrive before the first freeze in every state that has one. It is a planning anchor, and it is labelled as ours wherever it appears.
Five rodents, and they are not the same problem
House mouse
Mus musculus
The animal this page is mostly about. Small, grey brown, with a tail about as long as its body and a pointed nose. It is the only rodent for which an agency puts autumn entry in writing, and it is the one that fits through the smallest gap.
Norway rat
Rattus norvegicus
Heavy bodied, blunt nosed, with small ears and a tail shorter than its head and body. It burrows: under slabs, foundations, decking and woodpiles. Extension guidance for rats is about harborage, food and structural access, not about the calendar.
Roof rat
Rattus rattus
Slimmer than the Norway rat, with large ears, a pointed nose and a tail longer than its head and body. It climbs, and it nests high: attics, roof spaces, palms and dense ornamental shrubs. Its range is genuinely restricted, which is why this page checks it state by state.
Woodrat or packrat
Neotoma spp.
A native animal, not an introduced one, and a different problem with different rules. Woodrats build large stick middens, collect objects, and sometimes nest in parked vehicles. Several species are protected and one is federally endangered, so the published guidance leads with sealing rather than with control.
Deer mouse and white footed mouse
Peromyscus spp.
Native mice with white feet, a white belly and a sharply two-toned tail. They come into outbuildings, cabins and garages rather than into busy houses. They matter here for one reason: this is the group CDC names in connection with hantavirus, which is why the cleanup section on this page routes to CDC rather than to us.
The roof rat range, checked
The roof rat is the one species here whose range is genuinely restricted, and it is the one where the observation record can actually test the published range, because roof rats live in attics and fruit trees where people see and photograph them.
14 states have twenty or more research-grade records: Virginia, South Carolina, Georgia, Florida, Alabama, Mississippi, Louisiana, Nevada, Arizona, Texas, California, Oregon, Washington, Hawaii. 27 states have none at all, and they are the interior and northern states the published range describes as free of it. Two independent lines of evidence agreeing is worth something.
The classic published range puts the roof rat along the lower half of the East Coast, through the Gulf states and up into Arkansas, all along the Pacific Coast, and in Hawaii. Most of the interior of the country is described as free of it, with isolated infestations that probably arrived in cargo. Internet Center for Wildlife Damage Management, Roof Rats, from the 1994 handbook. Read it there.
The roof rat is described as less adaptable than the Norway rat and better suited to warm, tropical and subtropical climates, which is the stated reason it never spread across the whole country. Internet Center for Wildlife Damage Management, Roof Rats, from the 1994 handbook. Read it there.
California IPM states plainly that the documented range of the roof rat is expanding, without saying in which direction. UC Statewide IPM Program, Pest Notes: Rats, UC ANR Publication 74106, revised June 2025. Read it there.
And the direction that actually has evidence behind it. The documented expansion that does have a paper trail runs inland rather than north: roof rats were recorded in Phoenix in 2001, and by 2018 Arizona extension described them as well established across the state and reaching Yuma. University of Arizona Cooperative Extension, Roof Rats: Identification, Ecology, and Signs, az1775, July 2018. Read it there.
Where the range statements and the record disagree
Arizona extension says small roof rat populations exist in Alaska. There are no research-grade records of one in Alaska. And North Carolina and Arkansas both sit inside the published range yet have only eighteen and eleven records respectively. Both numbers are on the state pages.
Which is why the guidance for it leads with sealing. This is where the guidance genuinely diverges from the guidance for mice and introduced rats. Arizona extension states that native rodents are classified as protected nongame mammals under Arizona law, that several species in the desert southwest are endangered, and that the emphasis is therefore on sealing buildings and vehicle storage to keep the animals out. What you may lawfully do about a native woodrat is your state wildlife agency’s answer, so ask it before you act. University of Arizona School and Home IPM, Pack rats, July 2023. Read it there.
What to seal, and what not to bother with
Why exclusion and not something else. Exclusion is described as the most successful and permanent approach to house mice, and the working rule given is to close every gap and opening wider than a quarter of an inch. UC Statewide IPM Program, Pest Notes: House Mouse, UC ANR Publication 7483, revised July 2025. Read it there.
CDC, in its own words
“Fill small holes with steel wool. Put caulk around the steel wool to keep it in place or use spray foam. Use lath screen or metal, cement, hardware cloth, or metal sheeting to fix larger holes. Cut material to fit around pipes.”CDC, How to Seal Up to Prevent Rodents, page dated April 8, 2024 source
“Fix gaps in trailer skirting and use flashing around the base of the house. Seal outbuildings and garages as well.”CDC, How to Seal Up to Prevent Rodents, page dated April 8, 2024 source
“Move woodpiles 100 feet or more from your home and raise the wood at least one foot off the ground. Clean up brush and weeds around your home. Keep grass and shrubbery within 100 feet of your home well-trimmed.”CDC, How to Seal Up to Prevent Rodents, page dated April 8, 2024 source
“Keep bird feeders away from the house and use squirrel guards to limit access to the feeder by squirrels and other rodents.”CDC, How to Seal Up to Prevent Rodents, page dated April 8, 2024 source
What does not work
Plastic screening, rubber, vinyl, expanding foam, wood and anything else a mouse can gnaw are listed as ineffective for plugging holes, because the mouse simply chews back through. UC Statewide IPM Program, Pest Notes: House Mouse, UC ANR Publication 7483, revised July 2025. Read it there.
Steel wool pads are described as a good temporary plug that may rust over time, with metal or concrete named for the permanent repair around pipes, vents and cable entries. UC Statewide IPM Program, Pest Notes: House Mouse, UC ANR Publication 7483, revised July 2025. Read it there.
Lightweight chicken wire is ruled out for rat exclusion, because rats pass straight through openings that size. UC Statewide IPM Program, Pest Notes: Rats, UC ANR Publication 74106, revised June 2025. Read it there.
Mesh, and why one size does not do both. The federal wildlife damage handbook gives two mesh specifications rather than one: half inch by half inch woven hardware cloth in 19 gauge to exclude rats, and quarter inch by quarter inch in 24 gauge to exclude mice. A mesh sized for rats does not stop mice. Baker, Bodman and Timm, Rodent-Proof Construction and Exclusion Methods, in Prevention and Control of Wildlife Damage (USDA APHIS Animal Damage Control, Great Plains Agricultural Council and University of Nebraska), 1994. Read it there.
Before you staple mesh over a vent. The same handbook flags a tradeoff worth knowing before you staple anything over a vent: mesh finer than half an inch cuts air flow noticeably, and where ventilation is already marginal that can be unacceptable. In cold or dusty places fine mesh also clogs and freezes over. Baker, Bodman and Timm, Rodent-Proof Construction and Exclusion Methods, in Prevention and Control of Wildlife Damage (USDA APHIS Animal Damage Control, Great Plains Agricultural Council and University of Nebraska), 1994. Read it there.
The most productive gap in most houses. The gap between the bottom of an exterior door and its threshold is given a maximum of a quarter of an inch, with the note that building the threshold up is sometimes easier than modifying the door. Baker, Bodman and Timm, Rodent-Proof Construction and Exclusion Methods, in Prevention and Control of Wildlife Damage (USDA APHIS Animal Damage Control, Great Plains Agricultural Council and University of Nebraska), 1994. Read it there.
Two agencies disagree about foam
CDC names spray foam as a way to hold steel wool in place. California IPM lists expanding foam among the materials that do not work because a mouse gnaws through it. Both are published positions and they are incompatible. What both agree on is the permanent repair: metal, cement or concrete behind whatever goes in the hole.
Droppings: read CDC before you clean
This section is CDC, verbatim
CDC is a federal agency, so its material is in the public domain and is reproduced here exactly rather than paraphrased. Follow the links for the full guidance.
“Diseases are mainly spread to people from rodents when they breathe in contaminated air. Don’t vacuum or sweep rodent urine, droppings, or nesting materials. This can cause tiny droplets containing viruses to get into the air. If you already vacuumed, follow the cleanup guidance on this page.”CDC, How to Clean Up After Rodents, page dated April 8, 2024 source
“Open all doors and windows for 30 minutes before cleaning to ventilate the space. Leave the area during this time.”CDC, How to Clean Up After Rodents, page dated April 8, 2024 source
“Combine 1.5 cups of household bleach in 1 gallon of water (or 1 part bleach to 9 parts water). Make bleach solution fresh before use.”CDC, How to Clean Up After Rodents, page dated April 8, 2024 source
Which mouse, which disease
“The most common hantavirus that causes HPS in the United States is spread by the deer mouse.”CDC, About Hantavirus, page dated May 13, 2024 source
“Lymphocytic choriomeningitis, or LCM, is a disease caused by a virus and spread by the common house mouse, Mus musculus.”CDC, About Lymphocytic Choriomeningitis, page dated January 30, 2025 source
Two different animals, in two different parts of a property. The deer mouse is native and lives in sheds, cabins, garages and outbuildings; the house mouse is the one that comes into a kitchen. Blurring them is the most common accuracy failure on this subject, and it runs in both directions: it frightens people about the mouse in the pantry and it makes them casual about the shed.
“Rodent control is the primary strategy for preventing hantavirus pulmonary syndrome (HPS), a disease caused by infection with certain hantaviruses. Wild rodents near human populations should be controlled and excluded from homes.”CDC, Hantavirus Prevention, page dated May 13, 2024 source
How the season runs, nationally
August the month that matters
Nothing is happening yet, which is exactly the point. Alaska crosses 50F nights in the middle of the month, and the northern Plains, the northern Rockies and northern New England are only two or three weeks behind. If you live anywhere in that band, the caulk gun comes out now.
September the northern tier cools
Most of the Midwest, the interior West and New England cross 50F nights this month. Seattle, Portland and Denver too. This is the largest single month for cool down crossings in the country, and the point at which the outdoor food supply starts closing.
October the middle of the country
The Ohio Valley, the mid Atlantic, the mid South and the southern Plains cross now. Nationally this is also the month the raw house mouse observation record peaks, and where the raw and effort adjusted curves start agreeing with each other.
November the peak, and the South
The effort adjusted house mouse index peaks this month, and so does the roof rat index. The Deep South, Texas and the desert Southwest cross 50F nights now, weeks after everywhere else. Freezing nights arrive across the northern half and the influx there tails off.
December to March not seasonal any more
A population already inside a building is off the calendar entirely: heat, water and food mean it can breed through the winter. This is why sealing in November does not solve what arrived in September, and why the spring, once nights climb back above 50F, is the right time to do the work you did not get to.
Claims that circulate, checked against the agencies
Five claims, and what the sources say
- The "a mouse fits through a hole the size of a dime" line, and its "quarter for a rat" companion. It is not agency language, and the source people usually credit it to is CDC. We searched the full text of CDC’s seal up page and of its 2025 rodent brochure for dime, nickel, quarter, half dollar, penny and coin. Zero hits in either. CDC compares the hole to a pencil and gives a quarter of an inch. CDC publishes no rat hole size at all. We publish the inches.
- Long verbatim passages from the extension services, which is how most pages on this subject are written. Extension material is copyrighted even where the work is federally funded. UC ANR and UF/IFAS publish under noncommercial, no derivatives licences, and Ohio State, Minnesota, Kentucky, UMass and Mississippi State require written permission or bar commercial sites. CDC is federal and therefore public domain, which is why its wording appears here word for word while the extension findings appear in our own sentences with their names and links attached.
- Roof rats are marching north into the Midwest and the Northeast. Repeated constantly, and we could not trace it to any agency. The maps behind it are pest control industry service-data maps. The expansion that IS agency documented runs inland into the desert Southwest: Phoenix in 2001, Yuma by 2018, per University of Arizona az1775.
- A dollar figure for packrat damage to vehicle wiring. Every number we found came from a pest control company. University of Arizona says only that pack rats "occasionally" damage vehicles and wiring.
- The clinical detail on CDC's hantavirus pages, including the case fatality figure. It is CDC's own wording and it is accurate. It is also clinical rather than practical: it describes what happens to a patient, not what to do about a crawlspace. CDC's hantavirus pages carry it in full and are linked throughout this page.
Traps, bait and rodenticide
Control is a licensed trade in most states, and a licensed operator is the source for anything that goes in a bait station. The timing and the exclusion work are the parts a homeowner can do: the sealing date, the gap sizes and what the agencies say works. On rodenticide, a state wildlife agency describing what it finds in its own dead animals is a fact rather than a recommendation.
“Another very common way that wildlife is poisoned is by eating a rodent that ingested rodenticide (secondary poisoning). This occurs in species such as raptors, bobcats, and fox because rodents are a natural part of their daily diet. Poisoned rodents also become more lethargic and exhibit abnormal behavior such as sitting out in the open and making little to no attempts to escape from predators, and this increases the chance that they will be caught and eaten by a predator.”Oregon Department of Fish and Wildlife, Anticoagulant Rodenticide Fact Sheet, undated source
“One of the best things you can do is to prevent potential rodent problems. This includes plugging any holes that may provide entrance points in and out of buildings and reducing potential food sources or nesting areas.”Oregon Department of Fish and Wildlife, Anticoagulant Rodenticide Fact Sheet, undated source
Rodent season FAQ
Health information is quoted verbatim from CDC, a federal agency whose work is in the public domain, and linked. Secondary poisoning material is quoted from the Oregon Department of Fish and Wildlife. Findings from the cooperative extension services and from the federal wildlife damage handbook appear in our own sentences with their names and links attached, because several of those publishers restrict verbatim reuse on an advertising supported site; a fact is not copyrightable but the sentence expressing it is. The dates are ours, derived from NOAA NCEI 1991 to 2020 normal daily minimum temperatures, and are climatological typicals rather than forecasts. No agency publishes a sealing deadline; ours is labelled as ours everywhere it appears. Observation counts are research-grade records from iNaturalist, aggregate counts only, snapshot 2026-08-28. Related timing: fall home invaders and first frost dates.
Hero image: “House mouse (Mus musculus) 2732” — Ahunt — CC0