What the Federal Dashboard Says About Itself
The most honest description of the national West Nile virus numbers comes from the agency that publishes them.
On its current-year West Nile data page, the CDC states that ArboNET data are preliminary and subject to change, that current-season data are updated every one to two weeks from June through December, and that because of reporting delays, state, territorial, and local health departments may have more up-to-date information than what the federal page presents.
That is not a footnote. It is the agency telling readers, in plain language, that during an active season the federal dashboard is not the fastest or most complete source available.
ArboNET is run jointly by CDC and state health departments and covers human cases along with presumptive viremic blood donors, veterinary cases, mosquito pools, dead birds and sentinel animals. Its data currency at the time of writing is July 21, 2026.
For a disease with no vaccine and no specific antiviral, where prevention depends entirely on knowing whether transmission is occurring where you live, a one-to-two-week update cycle is a meaningful gap.
Five Limitations the CDC Publishes on Its Own Page
The agency lists specific caveats, and each one has a practical consequence for how a reader should use the numbers.
Underreporting comes first. Surveillance depends on a clinician considering West Nile as a possible diagnosis, ordering the right laboratory test, and reporting a confirmed result to public health. Every one of those steps can fail.
Mild illness is underreported more than severe illness. Non-neuroinvasive cases, meaning the febrile illness that does not involve the central nervous system, are far less likely to be captured, and CDC states directly that non-neuroinvasive surveillance data should not be used to compare disease activity between locations or over time. That single sentence rules out most of the comparisons readers instinctively want to make.
Cases are assigned by county of residence, not the county or state of exposure. Someone who lives in one county, works in another, and camps in a third is counted where they live. In a metro area, that scrambles the map.
Non-human surveillance varies by jurisdiction. CDC warns that absence of reported mosquito, bird or sentinel activity should not be read as absence of risk. A county with no reported positive mosquito pools may simply not be trapping and testing at the same intensity as its neighbor.
Finally, there is the reporting lag itself, compounded by states publishing on schedules that differ from CDC's.
What a County Dashboard Adds
Maricopa County, which has carried the largest share of this year's national West Nile burden, illustrates what local data provide that federal data cannot.
The county's public health department refreshes its dashboard weekly on Fridays, roughly twice the cadence of the federal update. It reported 29 confirmed human cases as of July 1, 2026, compared with 24 cases at the same point in 2025. Its full 2025 season total was 56 cases. The county has also disclosed that additional suspected cases were under review and not yet reflected in the confirmed count, which is a level of granularity the national page does not offer.
The county separately publishes mosquito surveillance results, listing traps and pools testing positive for West Nile virus and St. Louis encephalitis, with testing conducted by the Maricopa County Environmental Services Department. Positive mosquito pools are the leading indicator. They tell residents that infected mosquitoes are active in specific parts of the county now, weeks before any resulting human illness could be confirmed and reported upward.
That is the single strongest argument for checking local sources. Human case counts describe what already happened. Mosquito pool data describe what is happening.
Maricopa County also cautions that its current-year figures are provisional and can differ slightly from state-reported numbers because of differences in reporting timelines, which is the same honesty CDC applies to its own page.
Where the Numbers Disagree
Anyone comparing coverage of this season will notice that published death counts for Maricopa County have varied, with some reports citing four deaths as of early July and others citing five.
That discrepancy is worth naming rather than papering over, and it is exactly the kind of thing the reporting-lag structure produces. Death investigations close on their own timeline, outlets pull from the dashboard on different days, and a figure that was accurate on Tuesday may be superseded by Friday. Readers should treat any specific death count in secondary coverage as provisional and check the county dashboard directly.
The state-level distribution from the federal data as of June 30 showed Arizona with 32 cases, Texas with four, Colorado and Tennessee with two each, and one case each in California, Oklahoma, Nebraska, South Dakota, Arkansas, Florida and Pennsylvania. Because of the county-of-residence rule and the underreporting of mild illness, those single-case states should be read as evidence that transmission occurred somewhere, not as a precise measure of local risk.
What to Do with Better Data
Finding the right dashboard only matters if it changes behavior, and for West Nile it can.
Start with your county or state health department's mosquito-borne disease page rather than the national map. Look for positive mosquito pool reports in your area, spraying or larviciding notices, and the date the page was last updated. Many counties publish trap locations, which tells you whether your neighborhood is being monitored at all.
The prevention steps are unchanged and unglamorous because nothing else works. Use an EPA-registered repellent, cover skin during dusk and dawn when Culex mosquitoes bite most, repair window and door screens, and eliminate standing water in planters, buckets, drains, pool covers, and low spots in the yard. In monsoon regions, that last step needs repeating after every storm.
Risk is not evenly distributed. About one in five infected people develop a febrile illness, and roughly one in 150 develop neuroinvasive disease, which can cause encephalitis, meningitis or acute flaccid paralysis. Adults over 60, organ transplant recipients, and people with weakened immune systems account for the large majority of severe cases. Anyone in those groups who develops fever with severe headache, neck stiffness, confusion, muscle weakness or tremors after mosquito exposure should seek medical evaluation promptly.
August and September are historically the peak West Nile transmission months across most of the continental United States. CDC will continue updating ArboNET every one to two weeks through December. MedicalDaily will follow the national figures and county-level reporting through the peak.
Frequently Asked Questions
How often does CDC update West Nile case counts? Every one to two weeks from June through December. The current-year data were current as of July 21, 2026.
Why does CDC say local health departments have better data? Because of reporting lag. CDC states on its own page that state, territorial and local health departments may have more up-to-date information than the federal page shows.
What does county of residence mean for the map? Cases are counted where the patient lives, not where they were bitten. A case listed in one county may reflect exposure somewhere else entirely.
Why can't I compare mild case counts between states? CDC advises against it. Non-neuroinvasive illness is underreported at rates that vary by local awareness and health-seeking behavior, so the comparison is not valid.
Where should I look instead? Your county or state health department's mosquito-borne disease page, particularly positive mosquito pool reports, which are a leading indicator of local transmission.
Does no reported mosquito activity mean no risk? No. CDC specifically warns that absence of reported non-human activity should not be interpreted as absence of risk, because surveillance intensity varies.
Who is at highest risk of severe disease? Adults over 60, transplant recipients and immunocompromised people. Roughly one in 150 infections becomes neuroinvasive.
