Mass Killings

1. US mass killings by region, 2006–2025

The South accounted for 268 of the 631 recorded US mass killings in 2006–2025, or about 42%. The Midwest and West each had 149, while the Northeast had 65. The South therefore still contributes the largest share, as on the original page, but the updated two-decade series also shows that the two middle regions are tied. These totals include attacks in homes and public places, committed with firearms and other methods.

Raw regional shares reflect population as well as violence. The South has grown quickly and contains a large proportion of US residents; the Northeast is smaller and records the fewest incidents in this database. The chart does not show a resident’s risk and does not by itself establish which laws or social conditions explain the differences. Chart 6 uses population denominators for ten large states. Regions follow the US Census Bureau’s state assignments, and each incident is counted once in its recorded state.

Sources: AP/USA TODAY/Northeastern incident file, complete years 2006–2025; Census Bureau state and region classifications

2. Family, public and felony-related mass killings by region

Family mass killings are the largest category in all four regions. The South recorded 144 such incidents, compared with 67 in the West, 66 in the Midwest and 26 in the Northeast. The West had 36 public attacks and the South 43; the Northeast and Midwest had 19 and 22 respectively. This keeps the original chart’s emphasis on how the setting and relationship of victims can differ across the country.

The older commentary attributed regional differences to particular cultural traits and gun-ownership patterns, but this chart alone cannot test those explanations. The updated file identifies family and public attacks separately and also contains felony, suspected-felony and uncertain classifications. Here suspected felonies are grouped with felonies, while other, unsolved and undetermined cases form the final segment. These categories describe the incident circumstances recorded by the researchers; they should not be read as an assessment of motive in every case.

The Las Vegas shooting on 1 October 2017 remains the deadliest incident in this database, with 60 victims. Stephen Paddock fired from a hotel suite towards the Route 91 Harvest festival. The revised death toll includes two people who died later from their injuries; the image above is a view of Las Vegas retained from the original page, not a photograph of the attack.

Sources: AP/USA TODAY/Northeastern mass-killing incident classifications; Database scope and definitions, Northeastern University

3. Primary recorded method in US mass killings

Shootings were the primary recorded method in 499 of 631 mass killings during 2006–2025, about 79%. Stabbing accounted for 47 cases, fire or smoke inhalation for 41, blunt force for 27 and vehicle crashes for eight; nine used another primary method. The predominance of firearms is striking, but these figures also show why ‘mass killing’ and ‘mass shooting’ are not interchangeable. The broader category captures lethal attacks that involved no gun.

The original pie chart made a similar point over the shorter 2006–2017 period. This updated version uses the incident file’s first cause-of-death field, so a mixed-method attack appears under its first recorded method rather than being counted twice. The measure is incidents, not every individual victim’s cause of death. The 2017 Las Vegas attack, for example, is one shooting incident despite its exceptionally high death toll. Chart 11 separately follows the number of people killed each year.

Sources: AP/USA TODAY/Northeastern incident-level primary cause of death; AP database definition and firearm comparison

4. Method of mass killing by incident type

Firearms are the leading method in every incident type, especially public attacks: 110 of 120 public mass killings in this period were recorded primarily as shootings. Family killings were more varied, although shooting still accounted for 228 of 303. That group also included 34 stabbings, 16 blunt-force attacks and 15 cases led by fire. The chart extends the original breakdown while making the four groups directly comparable.

Felony-related killings included 110 shootings among 133 incidents, while the combined other or unknown group included 51 among 75. A stacked count emphasises how many events occurred in each type; it does not show the share within a category or the number of people who died. ‘Other methods’ pools stabbing, blunt force, vehicles and rarer means to keep the chart legible on a phone. Each event has one primary method even when a second cause of death was recorded.

Sources: AP/USA TODAY/Northeastern incident type and first cause-of-death fields

5. States with the most mass killings, by incident type

California recorded 69 mass killings in the completed-year series, followed by Texas with 51 and Illinois with 38. Florida had 34, Ohio 28 and New York 26. Family killings make up a large part of the total in most states, but public and felony-related attacks are also present. This retains the original chart’s state-and-type focus, with current classifications and twenty years of incidents instead of the earlier 2006–2017 snapshot.

Total counts partly follow population: a large state can have more incidents even when its per-person rate is lower. The graph therefore should not be used as a state safety ranking. A different view for the ten most populous states appears next, with incidents scaled by the resident population. The categories and state names come from the same AP/USA TODAY/Northeastern file, so the segments add exactly to each bar’s total. Small differences in a state tally may change when the research team updates individual case classifications.

Sources: AP/USA TODAY/Northeastern incidence file, state and type

6. Mass killings per million residents in ten populous states, 2021–2025

Illinois has the highest population-adjusted figure among these ten large states: eleven incidents in 2021–2025, equivalent to about 0.87 per million residents over the five years. Georgia follows at roughly 0.64 and Texas at 0.52. California recorded more incidents than any of these states, nineteen, yet its larger population puts its figure near 0.49 per million. The comparison shows why the raw state ranking above and a rate chart answer different questions.

These are cumulative five-year incident rates, calculated with the mean of each state’s 2021–2025 Census population estimates. They are not annual homicide rates or the proportion of people personally affected. Mass killings are comparatively rare, so a few additional events can move a state’s position sharply. The scope is the ten largest states by their 2025 estimated populations; it does not imply that smaller states have low rates or that state boundaries capture all differences in circumstances.

Sources: AP/USA TODAY/Northeastern incident dates and states; Census Bureau annual state population estimates, 2021–2025

7. School and college mass killings in the US, 2006–2025

The database records fourteen mass killings at a school or college in the twenty completed years from 2006 through 2025. Five occurred in 2011–2015, four in 2021–2025, three in 2006–2010 and two in 2016–2020. These are deeply consequential events, but the counts are small compared with the overall 631 mass killings in the same period. They also vary enough from one five-year span to the next that a simple rising trend is not apparent.

The original chart compared countries using a list of school massacres with five or more victims, mixing school shootings and other attacks without a consistently maintained cross-country dataset. This update keeps the school subject but uses the same verified US incident definition as the other charts: at least four people killed, excluding the attacker, within 24 hours, at a location coded school or college. Other violence linked to schools may fall outside that location rule, and many school shootings with fewer deaths are not counted.

Sources: AP/USA TODAY/Northeastern school or college location codes; AP database scope and definitions

8. US mass killings with at least 20 victims, by period

Seven mass killings in the AP series had at least twenty victims between 2006 and 2025. Four occurred in 2016–2020, including Las Vegas in 2017 with 60 people killed, Orlando in 2016 with 49 and El Paso in 2019 with 23. The other three were Virginia Tech in 2007, Sandy Hook in 2012 and Uvalde in 2022. All seven are coded as public killings in the incident file, so a type stack would contain only one segment.

The original page grouped historical US massacres since the nineteenth century and excluded massacres of Native Americans. That scope relied on a heterogeneous list whose categories and coverage could not be made consistent with a modern incident database. This chart keeps the question of extremely high-death-toll events but confines it to the research team’s verified 2006–2025 series. It says nothing about the prevalence of historical atrocities; the timeline begins in 2006 because that is when the source’s systematic tracking starts.

Sources: AP/USA TODAY/Northeastern victims killed per incident; Northeastern University database coverage

9. Largest US mass killings: victims, place, method and type

The incident table lists the 25 largest mass killings in this database between 2006 and 2025. Las Vegas in 2017 leads with 60 victims, followed by Orlando in 2016 with 49 and the Virginia Tech attack in 2007 with 32. Most of the very largest events were classified as public shootings. The table also includes the 2025 New Orleans vehicle attack, which illustrates why a page about mass killings should cover weapons other than firearms.

For readability, the on-page table is limited to the largest 25 incidents; the pack includes a CSV audit of all 631 completed-year cases, with region, city, state, victims, first method and type. Counts refer to victims killed and exclude perpetrators. The 2017 Las Vegas total is 60 because two people died later from their injuries. Rankings of the deadliest events should not be confused with the distribution of ordinary incidents: most cases in the database killed four or five people and took place in or near a residence.

Sources: AP/USA TODAY/Northeastern incident-level victims, place, cause and type; AP database presentation and definitions

10. US mass killings per year, 2006–2025

Annual counts move unevenly rather than steadily upwards. The highest completed year in this series was 2019, with 46 mass killings. There were 42 in both 2022 and 2023, then 38 in 2024 and 18 in 2025, the lowest recorded total since systematic tracking began in 2006. Earlier years also fluctuated: 38 incidents were recorded in 2006, while 2012 had 22. One year therefore offers only a partial view of the longer pattern.

The chart counts attacks in which at least four victims were killed within a 24-hour period, excluding the offender, by any method. It does not track the much broader category of shootings in which four people were shot but survived. The 2026 entries in the live source are omitted here because the year is incomplete as of this update. Incident totals may also be revised if the research team resolves an uncertain case. The fall through 2025 is substantial, but a few years do not establish a permanent trend.

Sources: AP/USA TODAY/Northeastern incident dates, completed years; Northeastern University database description

11. People killed in US mass killings per year

The annual number of people killed does not always follow the incident count exactly. An attack with a very high death toll can shift a whole year’s total: the Las Vegas shooting alone claimed 60 lives in 2017, while the Orlando attack claimed 49 in 2016. Across 2006–2025, the 631 mass killings in this file claimed 3,243 victims. Looking at victims alongside incidents shows both the frequency of these events and their sometimes very different human cost.

Each point is the sum of the incident file’s victim counts in that calendar year, excluding offenders. The series includes deaths from guns, knives, fire, vehicles and other means. It covers only completed years so that 2026 is not compared with full previous years. A chart of fatalities should not overshadow the people injured or the wider harms to families and communities, which are outside this particular measure. The two annual charts together make the difference between incident frequency and severity visible.

Sources: AP/USA TODAY/Northeastern victims killed by incident and date

12. Victims killed by incident type, 2006–2025

Family mass killings caused the largest combined loss of life in the dataset: 1,366 victims across 303 incidents. Public attacks killed 951 people in 120 incidents, fewer in total but considerably more per attack on average. Felony-related cases accounted for 592 victims and the other or unknown group for 334. The distinction helps explain why headline coverage often focuses on public attacks even though families account for the largest number of incidents and deaths overall.

These are people killed, rather than percentages of incidents. In the public category the mean is about 7.9 victims per event, compared with 4.5 in family killings, a difference heavily influenced by the largest attacks. Grouping suspected felonies with felonies and unresolved cases with other or unknown follows the same scheme used above. Victim relationships and circumstances can be complex, so these broad labels are a practical summary of the researchers’ classification rather than a full account of any individual tragedy.

Sources: AP/USA TODAY/Northeastern incident type and victims killed

13. Home and other locations over time

A residence or other shelter was the recorded location for 431 of the 631 mass killings in 2006–2025, about 68%. Homes therefore dominate the data across the four five-year periods, even though attacks in schools, workplaces, shops and public spaces are often more visible in national news. The pattern fits the large family category in the earlier charts, while also including some felony-related and other incidents at residential addresses.

Location and incident type are separate classifications: an attack in a home is not automatically coded as a family killing, and a public killing can occur at more than one type of place. Events spanning several settings are kept in the ‘other locations’ segment rather than counted in both groups. This chart focuses on where attacks happened, while the family and public charts focus on circumstances. The prevalence of residential attacks is one reason a complete discussion cannot be restricted to highly publicised school or venue shootings.

Sources: AP/USA TODAY/Northeastern incident location and date

14. How many victims were killed in a typical incident?

Most incidents sat close to the database’s four-victim threshold. The largest group involved exactly four people killed; the next group, five to nine victims, was also far more common than attacks causing double-digit deaths. Only seven events killed twenty or more people over the twenty completed years. This distribution puts the very large massacres in chart 8 into context without diminishing their impact on the victims, families and communities.

The threshold itself matters: cases with three victims are outside the AP database, regardless of how severe or widely reported they were. Counts here refer to victims killed, not injuries, and exclude the attacker. A different definition of a ‘mass shooting’ can produce a much higher incident total if it includes shootings with four people injured but fewer than four killed. Grouping by size shows what this specific series measures and why a few exceptional attacks can strongly affect the annual fatality line.

Sources: AP/USA TODAY/Northeastern victim count per incident; AP definition and comparison with other shooting counts