Flu season in the United States typically begins in October, with cases first rising in the South and Mountain West before spreading nationwide. Activity usually peaks between December and February, and the season can run as late as May. The precise start is not a fixed date but a threshold event: the week when influenza-like illness crosses the CDC's baseline and stays there. Most years, that crossing happens in late October or November.
That answer hides the interesting part. The virus does not consult a calendar. A collision of three forces — viral biology, indoor air physics, and human movement — decides when the season ignites each year. Understanding those forces explains why some seasons start in September and others smolder until January.
The biological trigger: why influenza prefers cold, dry air
Influenza A transmits through three routes: large respiratory droplets, fine aerosols, and contact with contaminated surfaces. Aerosols matter most, and aerosols are weather-sensitive. At low humidity, exhaled droplets evaporate quickly into small, light particles called droplet nuclei. Those nuclei stay airborne for hours and penetrate deep into the lower respiratory tract. At high humidity, droplets stay large, fall faster, and land on surfaces.
Temperature adds a second effect. Cooler air holds less moisture, so indoor heating in winter drops relative humidity further. Cold also stiffens the lipid envelope around the virus, which appears to keep it stable longer outside a host. The virus's own proteins work better at lower temperatures as well; studies of influenza A find that its surface proteins function more efficiently in cool, dry conditions.
The host changes too. Dry air thickens mucus and slows the cilia that sweep it out of the airways, giving the virus a longer window to attach. Seasonal changes in vitamin D, melatonin, and immune signaling may add smaller effects. None of these factors alone starts a season, but together they make late fall and winter permissive. Plainly: the virus travels better and the airway defends worse in dry, cold air.
Quantified context: what a season looks like in numbers
The CDC defines the start of elevated activity against a baseline built from pre-pandemic seasons. For recent seasons that baseline is roughly 2.5 to 3.0 percent of outpatient visits for influenza-like illness; when the weekly percentage exceeds the baseline for two consecutive weeks, circulation is considered elevated. In a typical pre-pandemic season, those weeks arrive in late October through November.
Case counts are far larger than lab confirmations suggest. The CDC estimates that influenza has caused roughly 9 million to 41 million illnesses, 140,000 to 710,000 hospitalizations, and 12,000 to 52,000 deaths annually in the US since 2010. Those ranges span a decade of seasons, and the variation is the point: the difference between a mild and severe season is often larger than the difference between an early and late one.
Timing has a measurable structure. In most seasons, hospitalization rates rise first in adults 65 and older and in children under 5, then broaden to working-age adults. Peak weekly hospitalization rates in recent pre-pandemic seasons fell between 10 and 30 per 100,000 people. From first elevation to peak is usually 6 to 10 weeks.
The timeline of a typical season
| Phase | Typical window | What the data show | Practical read |
|---|---|---|---|
| Seeding | September–October | Scattered cases, often among travelers and school-age children | Testing rarely captures spread |
| First elevation | Late October–November | Outpatient ILI crosses the 2.5–3.0 percent baseline | The season has started by definition |
| Acceleration | November–December | Weekly case and hospitalization counts climb steeply | Transmission is now sustained, not imported |
| Peak | December–February | Highest hospitalization and death rates; peak often late December to early February | Most seasons peak in this window but not all |
| Decline | February–March | Case counts fall but remain above baseline | Second smaller wave is possible |
| Tail | April–May | Activity returns below baseline; season declared over | Late-season cases still occur |
The table compresses large variation. Some seasons peak in November; others don't crest until March. The 2021–2022 season, for example, ran unusually late, peaking in April after a mild start. The shape is more reliable than the dates.
The regional paradox: why the South and Mountain West start earlier
Geographic timing within the US is strikingly uneven. The South — particularly Texas, Florida, and the Gulf states — routinely reports the first widespread activity and the earliest peaks, often in December. The Mountain West, including Colorado and Utah, shows a similar early curve. New England and the upper Midwest usually lag by four to eight weeks.
Humidity patterns explain part of it. Southern winters are milder but indoor heating and dry air can still produce low indoor humidity, and population density plus more indoor socializing in air-conditioned and heated spaces supports transmission. School calendars matter as well: early-starting school years concentrate children indoors sooner. Travel patterns to and from the tropics, where influenza circulates year-round, seed southern hubs earlier.
The counterintuitive detail: the region with the warmest winters often leads the nation in flu onset. Warmth outside doesn't protect people who spend their days in dry, recirculated indoor air.
Misconceptions that distort timing
Several beliefs routinely mislead people about when flu season starts.
"Flu season starts when it gets cold." Temperature is a contributor, not a trigger. Some humid summers produce outbreaks; some cold Decembers stay quiet until January. "The season always peaks in January." December-to-February is the statistically most common peak window, but peaks have occurred in November and as late as March. "Once you get the flu, you can't get it again that season." Multiple strains circulate, and immunity to one does not protect against another. "Vaccination timing is flexible because the season is long." Protection declines over months; early vaccination in August can leave you underprotected by a late peak. "Flu activity ends in spring." Strains can circulate through May and occasionally into summer, especially in the tropics.
The most consequential misconception is the vaccine-timing one. The CDC generally recommends vaccination by the end of October, but that is a target, not a deadline. Getting vaccinated in November or even later still provides protection for a late-starting season.
What determines an early versus a late start
Three variables dominate year-to-year timing. First, the dominant strain: influenza A(H3N2) seasons often start earlier and produce more severe outcomes in older adults, while A(H1N1) and influenza B seasons can run later. Second, population immunity: a season following a low-activity year tends to start earlier and spread faster because fewer people carry recent immunity. Third, school and travel patterns: school openings and holiday travel provide the mixing events that convert scattered cases into sustained transmission.
The 2020–2021 season illustrates the ceiling of these effects. Pandemic-era masking, remote schooling, and reduced travel drove influenza activity to record lows; the season essentially never crossed the baseline. When restrictions lifted, the following seasons arrived with unusual timing and severity. Plainly: human behavior can override the weather.
Practical actions, with thresholds
Timing knowledge is only useful if it changes behavior. Here are concrete steps.
Get vaccinated by late October if possible, and no later than the end of November. If you miss that window and activity is still rising, vaccination remains worthwhile. Watch the CDC's weekly FluView report. The season is officially elevated when outpatient ILI exceeds the baseline for two consecutive weeks; that is your signal to tighten precautions. Start hand hygiene and surface cleaning when local activity first rises, not at peak. Transmission chains are easiest to interrupt early. Consider masking in crowded indoor settings once hospitalization rates begin climbing, especially if you are over 65, pregnant, or immunocompromised. Seek testing within 48 hours of symptom onset if you are in a high-risk group; antiviral treatment works best early. Talk with a clinician about whether treatment is appropriate for you. Keep children home when febrile. Schools are primary amplification sites in the first weeks of a season.
Red flags that warrant prompt medical attention during any season include difficulty breathing, chest pain, confusion, dehydration, and fever that returns after improving. In infants, poor feeding and lethargy are warning signs.
Limitations: what timing forecasts cannot tell you
Season timing is a probability, not a prediction. The CDC and academic forecasters publish ensemble forecasts that estimate the probability of an early or late peak, but these models have wide uncertainty and perform worse in seasons with novel strains or unusual behavioral shifts. No model reliably predicts the exact week a season will start or how severe it will be.
Data also lag reality. FluView reports are typically one to two weeks behind actual transmission, and many cases are never tested. The baseline threshold itself is a statistical convention, not a biological switch. Two consecutive weeks above baseline reflects surveillance smoothing, not the moment transmission became sustainable.
For those reasons, treat "when does flu season start" as a range anchored to local data, not a date on a calendar. The most defensible rule: watch your region's weekly surveillance, expect elevation between late October and December, and prepare before it arrives.
This article is general educational information about influenza seasonality, not medical advice, a diagnosis, or a treatment plan. It does not promise any outcome from vaccination or other measures. For personal prevention and treatment decisions, consult a qualified local clinician, and seek emergency care for severe symptoms.