The CDC Again Cites Defensive Gun Use. What Do 60,000 to 2.5 Million Really Mean?

The Short Answer

The CDC currently points readers to literature estimating between 60,000 and 2.5 million defensive gun uses per year—approximately 164 to 6,849 incidents per day in national extrapolation.

Those are not cases counted by the CDC. They are endpoints associated with different surveys, definitions and screening methods. The CDC does not say that both endpoints are equally likely, endorse either one, or identify where within the range the true total lies.

One defensive gun use also does not automatically equal one life saved. An incident may concern property, a threat that never would have become lethal, disputed facts, or conduct later determined unlawful. But uncertainty is not evidence that the underlying events are fictional.

What matters to the responsible gun owner is the outcome. If danger ends, the intended victim survives, no innocent person is harmed and no shot is required, that peaceful ending is not an incomplete defensive use. It is the win.

What Changed—and Why It Is Newsworthy

A May 2022 archive included the numerical range. A later CDC-preserved September 2023 version said estimates varied but omitted the endpoints. The current CDC page, reviewed February 18, 2026, provides the range again.

Released correspondence and a Senate inquiry document advocacy directed at the CDC before the earlier deletion. The record establishes the advocacy and subsequent removal; it does not, by itself, establish the private motive of every CDC decision-maker. Restoring the language also does not resolve the scientific disagreement or convert 2.5 million into a CDC estimate.

The change is nevertheless worth noticing because public discussion often treats firearms only as sources of injury, detached from lawful ownership and protection. A firearm is a tool. It has no motive of its own, but its presence can magnify the consequences of human judgment in either direction. Responsible analysis must examine misuse, suicide and criminal violence without erasing lawful defensive use.

What the Research Actually Measures

The endpoints come from different research traditions, not one national database.

The lower-estimate tradition

The National Crime Victimization Survey begins with people who identify themselves as crime victims and then asks about actions taken during the incident. A 2024 peer-reviewed analysis of NCVS data from 1987 through 2021 estimated approximately 61,000 to 65,000 defensive incidents annually.

This approach limits some false positives by tying answers to reported victimization. But it can miss deterrence before an event is understood or reported as a crime, and respondents may be reluctant to disclose a gun-related encounter to a government interviewer. The study’s authors also emphasized the other side of the denominator: NCVS armed defenses were infrequent relative to the total volume of crime.

The higher-estimate tradition

Direct surveys ask a broader population whether they used a firearm for protection. The best-known is the 1995 Kleck and Gertz survey, which extrapolated approximately 2.2 million to 2.5 million annual incidents. Its annual estimate rested on 66 weighted positive respondents. Because defensive use is rare within a general-population sample, a few mistaken or misclassified answers can swing a national extrapolation substantially.

Higher estimates have been criticized for rare-event false positives, telescoping events into the recall period, social-desirability effects and uncertain legality. In separate surveys, five judges reviewing reported incidents rated a majority as probably illegal. That is a serious warning that direct surveys may capture questionable conduct as “defensive.” It does not prove that every high-end response was unlawful or invented.

A 2025 JAMA Network Open survey offers important contrary evidence. Among 3,000 adults with firearm access, 91.7% reported no lifetime defensive gun use and fewer than 1% reported any of the study’s four defensive behaviors in the previous year. The study did not produce a national annual incident estimate, so it cannot simply replace the CDC range. It does show why claims of millions of annual incidents need substantial qualification.

Why Estimates Differ So Much

The strongest objection to the low end is that victim-first screening and reluctance to disclose can omit real deterrence. The strongest objection to the high end is that rare false positives, memory errors and questionable classifications are magnified by extrapolation. Both criticisms can be partially true.

The upper endpoint deserves particular caution. The 2.5 million estimate remains part of the literature summarized by the National Academies and cited by the CDC, but it is driven heavily by a contested 1990s survey in which a small number of positive responses produced a very large extrapolation. Later direct surveys face related problems. This does not establish that the true number is near zero; it means the upper bound is a disputed endpoint, not a likely national count.

The wide range should therefore produce humility—not dismissal. Even the low-end literature estimate represents roughly 164 incidents per day in national extrapolation. That is substantial without pretending it is a count of verified police cases or saved lives.

Why a Peaceful Ending Is a Win—and Often Leaves Almost Nothing to Count

No wound and no body create very little administrative evidence—and that is a good thing.

A responsible gun owner does not carry because he or she wants to pull the trigger. The firearm is insurance against the moment when avoidance fails and a dangerous disparity of force leaves no safe alternative.

The simple win is survival: the intended victim is safe, innocent people are unharmed, no shot is fired if it can be avoided, force stops when the threat stops, and even the aggressor remains alive when circumstances allow. The goal is not to win a gunfight. It is to prevent one, escape one, or survive one with the least harm reasonably possible.

If an aggressor recognizes that the intended victim can defend herself and leaves, there may be no discharge, injury, ambulance, hospital record, recovered bullet, news story or police report. The responsible carrier may be satisfied that everyone went home alive.

In the 2021 National Firearms Survey working paper, 81.9% of respondents’ reported defensive incidents involved no shot. That self-reported result is not an independently verified national rate and does not establish that every incident was lawful. It illustrates the measurement problem: many claimed encounters end through warning, recognition or withdrawal rather than gunfire. A peaceful ending may create little administrative evidence precisely because no one was injured.

Quiet records cannot prove that the real number exceeds 2.5 million. They explain why a complete administrative count is probably impossible and why “no documented injury” should not automatically mean “nothing happened.” We should never inflate the data—but we should not define a peaceful outcome out of existence because it produced no wound to count.

Why This Matters Beyond the Statistics

Americans commonly possess and carry firearms for lawful purposes, including personal protection. In Heller, the Supreme Court recognized individual self-defense as central to the Second Amendment. In Bruen, it reaffirmed the right of ordinary citizens to carry a handgun publicly for self-defense and noted that handguns are in common use for that purpose.

The CDC range does not decide a constitutional case, establish that every reported incident was lawful, prove how commonly a particular firearm or accessory is possessed, or answer whether a regulation fits the nation’s historical tradition.

Heller (2008), Bruen (2022) and Rahimi (2024) treat lawful self-defense as central to the Second Amendment right, but none turns on whether the national total is 60,000 or 2.5 million. The statistics inform public debate; they are not the constitutional test.

The totals do not create the right. They provide factual context for a lawful protective purpose already recognized by the Court. Public-policy discussions should not treat firearms solely as instruments of injury while ignoring that protective role.

The Strongest Objections—Answered Honestly

The Responsible New York Takeaway

Constitutional protection is not permission to threaten, display or use a firearm whenever someone feels afraid or angry. New York’s justification statutes distinguish ordinary physical force from deadly physical force and impose strict limits. A pistol license is not authorization to brandish, punish, escalate or pursue.

• Create distance and leave when safely possible.
• Do not allow pride or anger to turn an argument into a force encounter.
• Never assume displaying a firearm is a harmless warning.
• Use no more force than the law and immediate circumstances justify.
• Stop when the threat stops.
• Seek qualified legal advice about a specific incident.

Continue with NY Safe Inc.’s guidance on concealed-carry judgment, New York use-of-force law, disparity of force, and the legal aftermath of a defensive shooting.

Training Is About Judgment Before It Is About Marksmanship

NY Safe Inc. teaches lawful, responsible concealed carry with an emphasis on avoidance, decision-making, safe handling and the realities following a defensive encounter.

Watch the Free “Should I Carry?” Webinar

Frequently Asked Questions

Sources and Method Notes

External links open in a new tab. Estimates are rounded and should be read with each source’s definitions and limitations.

• CDC: About Firearm Injury and Violence Prevention
• Archived CDC page, May 2, 2022
• CDC-preserved page, September 2023
• National Academies research priorities report (2013)
• National Academies critical review (2005)
• McDowall and colleagues, NCVS trends (2024)
• Kleck and Gertz (1995)
• Hemenway and colleagues, probable legality review
• CDC BRFSS historical questionnaires
• Kleck, BRFSS reanalysis (2020)
• English, 2021 National Firearms Survey working paper
• JAMA Network Open survey study (2025)
• Senate inquiry and attached correspondence
• New York Penal Law Article 35

Legal notice: This is general educational information, not legal advice. The legality of defensive conduct depends on precise facts and applicable law. Consult a qualified New York attorney about a specific situation.