Preliminary 2025 U.S. Mortality: Death Rates Down Again!
Great News for U.S. Birthday! Let's Celebrate Good Times!
These are preliminary data from the CDC, but this accords with what I had been expecting… and I should be able to do a good preliminary ranking table by cause of death next week, by age group and by sex.
I can update a few key graphs given the data from the CDC right now, though.
2 Jul 2026, CDC: Provisional Data: U.S. Death Rate Fell to a Record Low in 2025
NCHS has released a new report—”Mortality in the United States: Provisional Data, 2025” that provides early estimates of deaths that occurred in the United States last year. According to the report, an estimated 3,094,593 people died in the United States in 2025. The death rate was 689.2 per 100,000—a 4.6% decrease from 2024 and a record low.
The new report includes numbers and rates of death by age, sex, and race and Hispanic origin. Rates fell for every age group and for both men and women in 2025.
The report also lists the 10 leading causes of death in 2025. The top three estimated causes of death in 2025 were heart disease with 694,708 deaths, cancer with 622,832 deaths, and unintentional injuries with 184,265 deaths. The report also notes changes in other cause rankings: influenza and pneumonia rose to the eighth leading cause of death, and suicide dropped from the tenth to the eleventh leading cause.
Let’s check out this report!
U.S. Preliminary Death Rates for 2025: Crude and Age-Adjusted
July 2026, NVSS: Mortality in the United States: Provisional Data, 2025
Results—In 2025, 3,094,593 deaths occurred in the United States. The age-adjusted death rate per 100,000 population was 811.1 for males and 582.9 for females. The overall rate, 689.2 per 100,000 population, was 4.6% lower than in 2024. The rate decreased for all age groups. Overall, age-adjusted death rates in 2025 were lowest for the multiracial non-Hispanic population (187.3 per 100,000 population) and highest for the Black non-Hispanic population (869.0). The leading causes of death were heart disease, cancer, and unintentional injuries.
I will note very quickly that the numbers above do not match what is currently sitting in CDC WONDER…. because CDC WONDER currently has last year’s population estimates sitting there.
An age-adjusted death rate of 725.5 and 689.2 are 5% different… if you use the Census estimates that were released in June: National Population by Characteristics: 2020-2025, even without using all the adjustments done by the CDC (they don’t use the mid-year estimates directly), I was able to calculate 689.4 as an age-adjusted rate using the deaths currently reported by WONDER.
If you want to see how to calculate age-adjusted death rates, check out this post:
Good enough for me to proceed with further analysis.
Most of the Improvement from Teens/Young Adults
Here’s the good news: all that excess mortality that I’ve been talking about in teens and young adults through the pandemic from homicides, motor vehicle accidents, and drug overdoses?
While I haven’t looked at cause of death data yet, first, let’s look at the age groups:
They all decreased in death rates between 2024 and 2025. Huzzah!
But I highlighted the ages 15-44 — these were our “pain” ages before, especially for drug overdose deaths.
When it comes to age-adjusted death rates, not all age groups are created equal — the oldest age groups are weighted less than prime working ages, for instance. It’s based on a “standard” population age distribution (based on the population age distribution near the year 2000… we’ll be re-calibrating to 2030 … soon… soon…)
So which ages had the biggest impact on the change in the age-adjusted death rate?
Exactly.
The large improvement combined with the larger weights on ages 15-44 led to those having the largest impact. 65% of the effect came from that age range.
Longer-term trends, 1968-2025
The age-adjusted rate is indeed lower than ever before.
The crude rate (number of deaths over total population) is high… because we’ve got a bunch of older folks, and the population is likely to become even more biased toward a larger percentage of the population being at older ages. They do tend to die at higher rates. This is why we do the age-adjustment.
It will be interesting to see how the 2030 age-adjustment differs from the 2000 age-adjustment … I have already seen how the 1970 and 1940 weights give different results for certain periods.
From 4 years ago:
