Infectious disease and vaccination, state by state

About this data

This map brings together more than 100 years of public records. Here is how we checked it, and what to know before using the numbers. The full method, source by source, is in DATASETS.md.

How we checked

• Every figure is built from the original tables and files listed under Sources above, by code that is in the project repository.
• Before publishing, an independent check rebuilt the numbers from those files without using ours. On October 15, 2026 it recomputed about 12,000 state-year case counts, every state population and every vaccination figure, and reread 177 numbers by eye from scanned government reports. None differed from the map.
• On November 11, 2026, a second independent check covered what was added after that: the 2026 case counts through November 7, the 2026 rates on 2025 population, the 2026 county measles counts and the 2025-26 kindergarten figures, along with the wording on the page. None of its numbers differed from the map.
• Neither check covered the state facts on each state's page (income, poverty, schooling, health insurance, jobs and party control) or the headlines above the timeline.

Case counts

• Reported cases are not all cases. Many people who got sick were never counted, and the rules for what to report changed many times.
• Before 1993 the counts come from scanned government reports, except Alaska and Hawaii before statehood. A few printed numbers could not be read with certainty: the 1981 measles page curves at its edge, so two neighboring states in one region could be swapped; one digit in Arkansas's 1958 whooping cough count is damaged (we use 156, the value that makes the regional total add up); and Arizona's 1975 and 1976 mumps and whooping cough print a dash, shown as zero, between years marked as not reportable.
• In the 1920s the yearly tables print a dash for some states. The tables define a dash as zero, so the map shows zero, but some of these states may not have been reporting the disease yet. For example, Virginia whooping cough is 0 in 1922 and 1923, then 17,710 in 1924.
• Some printed counts include other diseases, and the table says so in a footnote: rubella in some measles counts (Nebraska 1955 to 1958, Mississippi 1957 and 1967 to 1969, South Carolina 1964), and other types of hepatitis in some hepatitis A counts (Louisiana 1966 to 1969, Georgia 1976, Tennessee 1977, New York 1977 to 1984). The hover box for each one says so.
• Hepatitis A before 1974 includes hepatitis of unspecified type. CDC revised the hepatitis A case definition in January 2019, and from 2024 its weekly tables count confirmed cases only, which is a narrower count.
• Polio counts all reported polio before 1956 and paralytic polio only from 1956.
• 2025 and 2026 counts are provisional. Final counts are usually higher. 2026 is not over: it covers CDC's weekly reports through , and its rates use 2025 population because there is no 2026 estimate yet. Because the year is not finished, a 2026 rate runs lower than a full year would, so a state near a line between two colors may show one step lower.
• Alaska (through 1958) and Hawaii (through 1959) come from Project Tycho's copies of the weekly reports: sums of the weekly counts, or for polio from 1956 to 1959 the year-end running total. These usually run lower than a full year's count. In those years the United States row adds them in, so it runs above CDC's printed national totals (measles in 1956 by 7,665 cases). For polio from 1956 to 1959, it adds their counts of all polio to a national count of paralytic polio.
• New York City and the rest of New York were printed separately in some years. Where only New York City was printed, New York shows no data for hepatitis A in 1985 and 1986, and the United States row leaves those cases out. Mumps from 1968 to 1973 is New York City only, shown with no rate.

Vaccination

• Each state reports kindergarten vaccination in its own way, and some changed their method over time: 9 of the 46 states with figures in both years counted differently in 2024-25 than in 2019-20. Comparisons between states should be made with care.
• CDC marks some states' figures as lower bounds, shown here as “at least.” It marks others as counting doses given at any age or spacing, which may make them higher than the share with valid doses (12 states in 2024-25). The hover box lists CDC's notes for each state and year.
• The toddler survey changed how it reached families: landline phones only through 2010, landline and cell phones from 2011, and cell phones only from 2018. CDC says figures from 2012 on are not directly comparable with earlier years. In 2019 CDC also changed from reporting by survey year to reporting by birth year, and the charts leave a gap between the two.
• National figures before 1995 come from surveys that took parents' word without checking records. The survey's authors say this understated coverage, by as much as 23%. A “don't know” answer added in 1976 breaks the series there. The survey and age group changed in 1991 (children 1 to 4 before 1986, toddlers from 1991), and there is no national figure for 1986 to 1990. For 1959 to 1969 and 1977 to 1982 we found only one source. For polio in 1970, two sources give 65.9% and a third gives 77.5%; the map uses 65.9%.

County view

• County vaccination figures were compiled by Johns Hopkins University from state reports. Most states count kindergartners, but some count other grades, so compare a county with its own state rather than across state lines.
• Wisconsin caps its county figures at 95%. Johns Hopkins notes that Wisconsin moved from registry data to a school survey and that its earlier years were updated to the new data. Figures Johns Hopkins marks as not comparable (Utah 2017-18, Missouri 2019-20, Louisiana 2020-21 and 2021-22) are left out.
• County measles cases come from the Johns Hopkins measles tracker, which counts from different reports than CDC, so its totals are not the same as CDC's. Its 2026 cases run through .