Violent Deaths in the United States, 2021: What the National Violent Death Reporting System Found

In 2021, about 75,000 people in the United States died of violence-related injuries. CDC's National Violent Death Reporting System (NVDRS) goes beyond the death certificate: it pulls together death certificates, coroner and medical examiner records, and law enforcement reports, records the circumstances that led up to each death, and links related deaths — several homicides, or a homicide followed by a suicide — into a single incident.

CDC

Coverage

NVDRS began in 2003 with six states — Maryland, Massachusetts, New Jersey, Oregon, South Carolina and Virginia — and has grown since. CDC has funded all 50 states, DC and Puerto Rico since 2018.

This report covers 48 states — every state except Florida and Hawaii, whose data did not meet the completeness threshold for circumstances — plus DC and Puerto Rico. Forty-six states reported statewide; California reported from 31 counties and Texas from 13. The 48 states and DC were 85.4% of the U.S. population and 86.5% of its violent deaths.

States and jurisdictions by the year they began collecting NVDRS data, 2003–2021. CDC

2021 in 48 states and DC

70,688 deaths in 68,866 incidents:

59.2% of all these deaths involved a firearm, and 60.4% of victims were injured in a house or apartment.

"Legal intervention" means a death caused by law enforcement or others with legal authority to use deadly force, acting in the line of duty — legal executions excluded. It is an ICD-10 classification and says nothing about whether the death was lawful.

Suicide

White people accounted for 77.1% of suicides. Among women, those aged 45–54 had the highest rate (8.7).

Method. A firearm was used in 54.3% of suicides, then hanging, strangulation or suffocation (26.2%) and poisoning (10.7%). A firearm was the commonest method for both men (59.5%) and women (33.7%). 71.3% happened in a house or apartment.

What came before (known for 84.4%):

About half had no known mental health diagnosis, and fewer than one in four were known to be in treatment — a gap, the authors say, between who gets treatment and who would likely benefit. Women were more likely than men to have a diagnosed condition (64.9% vs 44.9%) and to be in treatment (35.7% vs 19.6%).

Of those tested, 40.2% had alcohol in their blood, and 65.1% of them were at or above 0.08 g/dL; 22.2% tested positive for opioids and 35.7% for antidepressants.

Homicide

Black people were 61.1% of male and 43.5% of female homicide victims.

Method and place. Firearms were used in 79.9% of homicides, then a sharp instrument (8.1%) and a blunt instrument (3.2%). The firearm homicide rate for males was 5.4 times that for females. Women were far more often killed at home (a house or apartment: 59.7% vs 35.8%); men more often on a street or highway (24.7% vs 10.6%).

Who and why. The relationship to the suspect was known in only 32.7% of homicides. When it was:

• men were most often killed by an acquaintance or friend (29.1%) or someone else they knew (22.9%), then a stranger (18.1%);
• women were killed by a current or former intimate partner in about half (51.0%) of cases.

Circumstances were known in 70.3%. The commonest were an argument or conflict (33.5%) and another crime (21.9%, mostly assault or homicide, then robbery). Intimate partner violence precipitated 41.5% of homicides of women but 7.2% of men — and where it did, women were usually the partner themselves, while men were more often bystanders to it, such as a new partner killed by an ex.

Legal intervention deaths

94.5% were male. The highest rate was among men aged 30–34 (1.7 per 100,000). AI/AN males had the highest rate (1.9), 3.8 times White males (0.5); the rate for Black males (1.4) was 2.8 times White males'. A firearm was used in 86.8%. Where circumstances were known, the person who died had used a weapon in 66.8%, had a substance use problem other than alcohol in 24.9%, and a diagnosed mental health problem in 19.9%. 29.6% were in custody when injured.

Unintentional firearm deaths

Of 555, 43.1% were self-inflicted and 36.4% inflicted by someone else (unknown for 20.5%). 84.3% were male, and 56.0% were 24 or younger. 73.3% happened in a house or apartment; 61.4% involved a handgun. A child was present or saw it in 27.4%.

23.7% followed an unintentional pull of the trigger, 20.1% someone thinking the gun was unloaded, and 7.8% the gun being mistaken for a toy — which, the authors say, shows why secure storage and education on safe handling matter.

Deaths of undetermined intent

Rates were higher among males, especially Black (5.9 per 100,000) and AI/AN males (5.4), and highest among men aged 30–34. Poisoning was the method in 65.7%. Of those tested, 77.0% were positive for opioids, 50.9% for antidepressants and 43.2% for cocaine; 70.1% of those with known circumstances had a substance use problem other than alcohol.

Puerto Rico

Puerto Rico's 880 deaths in 816 incidents were analyzed separately. There, the homicide rate was more than double the suicide rate:

• Homicide: 602 men and 37 women — a male rate 17.7 times the female rate. Among men aged 18–29 the rate was 93.7 per 100,000. 91.5% involved a firearm. Among men, 59.2% of homicides were gang related and 30.4% involved drugs; intimate partner violence precipitated 41.7% of homicides of women.
• Suicide: 186 men and 29 women. Hanging, strangulation or suffocation was the commonest method (64.5% of men, 55.2% of women); a firearm was used in 18.3% of men's suicides.

What helps

Violence is preventable. CDC's resources for action gather the best available evidence on preventing suicide, youth violence, child abuse and neglect, adverse childhood experiences, and intimate partner and sexual violence. Its suicide resource names seven strategies:

1. strengthen economic supports;
2. create protective environments — including reducing access to lethal means for people at risk;
3. improve access to and delivery of suicide care;
4. promote healthy connections;
5. teach coping and problem-solving skills;
6. identify and support people at risk;
7. lessen harms and prevent future risk.

Its child abuse and neglect resource lists strengthening economic supports for families, social norms that support parents and positive parenting, quality care and education early in life, parenting skills, and intervening to lessen harms and prevent future risk.

States already use their data this way:

• North Carolina's VDRS data were a primary source for partners and led to the state's Office of Violence Prevention, focused on firearm deaths.
• Maine's data on law enforcement officer suicides helped support mandatory mental health resiliency and awareness training at the Maine Criminal Justice Academy, with similar training planned for corrections officers.
• Georgia's VDRS worked with the City of Atlanta Mayor's Office of Violence Reduction on two public dashboards showing where people at higher risk of homicide live.

Limits

• Not nationally representative: 48 states, DC and Puerto Rico, with only some counties in California and Texas. Hawaii's absence means the figures for Native Hawaiian and other Pacific Islander people may be incomplete.
• Data depend on partnerships with coroners, medical examiners and police, and can be thin where those are fragmented.
• Toxicology isn't done for everyone or consistently.
• Abstractors see only the reports they receive; homicide and legal intervention records may be incomplete until investigations close.
• Suicides of children 5–9 are not included.
• Mental health information often comes from family and friends, not medical records.

Coming next

A School-Associated Violent Death Module (deaths from January 1, 2021) and a Public Safety Officer Suicide Module (from January 1, 2022) are being added. NVDRS moved to a secure cloud system in 2023. The goal is data from all 50 states in future reports.

If you or someone you know is struggling or in crisis, call or text 988, or chat at 988lifeline.org, to reach the 988 Suicide & Crisis Lifeline.

Sources

Based on Nguyen BL, Lyons BH, Forsberg K, Wilson RF, Liu GS, Betz CJ, Blair JM, "Surveillance for Violent Deaths — National Violent Death Reporting System, 48 States, the District of Columbia, and Puerto Rico, 2021," MMWR Surveillance Summaries volume 73, number 5, Centers for Disease Control and Prevention; a work of the United States government in the public domain. CDC notes the report has been corrected. Where the report gives different figures for the same thing, the figure is left out: the rate of deaths of undetermined intent (2.0 and 2.1 per 100,000), and the share of California's population the 31 counties represent (64% and 62%). Where the discussion rounds a result, the result itself is used.