Black maternal mortality is 44.8 per 100,000 — more than three times the White rate of 14.2

What is happening?

In 2024 the United States recorded 649 maternal deaths, a rate of 17.9 per 100,000 live births, against 669 deaths and 18.6 in 2023. Inside that national figure the rate for Black non-Hispanic women was 44.8 — set against 14.2 for White non-Hispanic women, 18.1 for Asian non-Hispanic women and 12.1 for Hispanic women. NCHS states that the 2024 rates were not statistically different from 2023 overall, by age, or by race and Hispanic-origin group, so the gap did not measurably move in the most recent year on record.

The review side of the same question reaches a sharper conclusion. Committees that read the case files of pregnancy-related deaths judged 87 percent of them preventable in the most recent CDC release, which covers 2021 deaths reported by 46 states and was published on 2025-08-22. Discrimination was recorded as a contributing circumstance in 32 percent of those deaths.

Whose problem is this?

The party that measures the deaths and the party that produces them are not the same party, and neither of them is the party that pays. CDC counts and publishes; hospitals and clinicians deliver care; Congress and the states decide whether the counting and the coverage continue. A committee finding of preventable lands on no ledger.

Where does this problem end?

The boundary matters because the national rate and the disparity are two different problems with two different remedies. A policy that lowers the national rate by improving care for the majority group can leave the ratio exactly where it was, so the two numbers have to be read separately.

What is the state now, and what should it be?

Now

The country already produces a rate of 14.2 for one group and 44.8 for another in the same year, under the same clinical guidelines, inside the same hospital systems.

Needs a new measurementThe target state: none of the sources opened for this file states an official federal or national target for the size of the gap, or a date by which the Black rate should reach any particular value. What the sources establish is the comparison itself, not a goal.

How big is it?

The population carried here is the annual cohort exposed to the elevated rate, not the count of deaths.

Of the 3,628,934 live births registered in the United States in 2024, 13.2 percent were to non-Hispanic Black mothers, which works out to about 479,019 births. That is the lower bound. The upper bound of 515,308 adds the 0.7 percent of 2024 births to American Indian and Alaska Native mothers and the 0.3 percent to Native Hawaiian and Pacific Islander mothers — two groups whose 2024 maternal mortality rates NCHS did not publish because the counts are too small to release, and whose 2021 pregnancy-related mortality ratios of 118.7 and 111.7 per 100,000 were the highest of any group.

Two arithmetic consequences follow from the published rates. At 44.8 per 100,000, the 479,019 Black non-Hispanic births imply roughly 215 maternal deaths — about a third of the national 649, from 13.2 percent of the births. Had that cohort died at the White non-Hispanic rate of 14.2 instead, the implied count would be about 68, so the gap is worth roughly 147 deaths a year. Both figures are derived from rates rather than read off a published count of deaths by race.

Under what conditions does it arise?

None of these conditions is a shortage of medical knowledge about how to keep a pregnant woman alive. Every one of them is about who is reached, by which program, in which state, and for how long after the birth.

What has been tried?

The attempts split cleanly into three kinds — measure it, pay for it longer, and publish it. None of the five is an intervention at the bedside, and the opened sources do not report whether any of them changed the ratio.

What was found?

Why is it still unsolved?

Supply disparity — the care that produces a rate of 14.2 exists inside this country and is not reaching the people who need it most.

The service exists in this country, it demonstrably works, and it does not arrive evenly. That is what separates this problem from one of research or of technology. When a review committee reads a case file and writes down preventable, it is saying that the knowledge, the drug, the transfusion protocol or the follow-up visit was available somewhere in the same health system on the same day. Eighty-seven percent of the time, that is the finding. A problem in which the answer is known and the answer does not arrive is a distribution problem.

The distribution runs through channels that no single actor controls. More than two-thirds of births to Black women are financed by Medicaid, and Medicaid is fifty separate programs with fifty sets of rules. The extension of postpartum coverage from 60 days to 12 months reached 48 states and DC only by 2026-02, and one state has still not taken it. Place compounds the effect independently of payment: review-committee data put rural mortality ratios over 50 percent higher than in large urban areas, and the spread across states with publishable rates runs from 10.1 to 42.1.

The counting is contingent in a way the counted are not. The national picture of preventability comes from one system, funded through one federal program, staffed by one CDC division. That program was reauthorized in 2026-02 through 2030, and in the year before that a departmental reduction in force disrupted parts of the CDC capacity that runs it. A gap that is visible only because someone chose to publish it by race can become invisible without the underlying rate changing at all.

What observation would mean it is solved?

Candidates — (a) the ratio of the Black rate to the White rate falls and stays below some agreed threshold, (b) the Black non-Hispanic rate itself falls to the level currently recorded for White non-Hispanic women, (c) the share of pregnancy-related deaths judged preventable by review committees falls sharply.

No single number can carry this, because each of the three candidates can improve for a reason that has nothing to do with the gap closing.

• (a) is a ratio, and a ratio improves when the denominator worsens. If the White rate rose to 20 while the Black rate held at 44.8, the ratio would fall from 3.15 to 2.24 and nothing good would have happened.
• (b) is the honest target and it is the slowest to read. With 649 maternal deaths nationally and an implied 215 in this cohort, a single year of movement is inside the noise — NCHS itself reports that no 2024 rate differed significantly from 2023.
• (c) can fall because review capacity fell. Preventability is a committee judgment over cases the committee sees. Fewer committees, thinner staffing or slower reporting all push that share around without moving a single death.

A fourth trap sits under all three: the denominator is shrinking. The Black non-Hispanic share of US births fell from 14.3 percent in 2016 to 13.2 percent in 2024, so the absolute count of deaths in this cohort can decline while the rate does not.

What is it connected to?

Fills with researchthe closure of rural maternity units, Medicaid payment policy, severe maternal morbidity, and the wider question of how US maternal outcomes compare with other high-income countries all plausibly touch this problem. Relation type and evidence grade were not established in this round.

What these sources do not say

• State rates broken down by race. The state-level table covers 39 states as a 2019–2023 aggregate and is not split by race in any source opened here, so the place where the gap is widest is unknown from this file.
• The count of Black maternal deaths. Every death figure here is either national or derived from a published rate. No source opened gives the number of Black non-Hispanic maternal deaths in 2024 directly, so the roughly 215 above is arithmetic, not a count.
• Whether any intervention moved the ratio. Five attempts are listed and not one of the sources reports a measured effect of any of them on the gap. Coverage converged to 48 states plus DC and no opened source connects that to the 2024 rates.
• What discrimination means in the case files. It is recorded as a contributing circumstance in 32 percent of pregnancy-related deaths, but the sources opened do not say how a committee decides that, who decides it, or how consistently the judgment is applied across 46 states.
• Why the two measures disagree. NCHS puts Asian non-Hispanic women above White non-Hispanic women in 2024 at 18.1 against 14.2, while the review-based 2023 figures put them below at 10.7 against 14.9. The definitions differ in window and in system, and no source opened reconciles them.
• The 2024 rates for American Indian and Alaska Native and for Native Hawaiian and Pacific Islander women. Their 2021 pregnancy-related ratios were the highest of any group, and their 2024 rates are absent because the counts fall below the release threshold. The groups with the worst recorded outcome are the ones the annual series cannot report.

See the evidence

The NCHS primary document was not read directly. Every attempt to fetch cdc.gov returned HTTP 403, so the 2024 figures here come from the PubMed record of Health E-Stat 113 and from two independent reports of the same 2026-03-04 release. Those two reports agree with each other and with the PubMed abstract on every value they share — 17.9 for 2024, 18.6 for 2023, and 44.8 for Black non-Hispanic women — and the total births figure of 3,628,934 was confirmed twice, in the final natality report and in a separate NCHS data brief. One date is unresolved: the PubMed record carries September 2024 in its NCHS Health E Stats series line while the release itself is dated 2026-03-04 by the news coverage, and no source opened here explains that difference.

Two sets of numbers on this page describe different things and should not be read as a disagreement. The 44.8 and 14.2 are maternal mortality rates counting deaths within 42 days of the end of pregnancy, from death certificates. The 49.4 and 14.9 are pregnancy-related mortality figures counting deaths within one year, from review committees. They come from different systems with different windows, and the ordering of the Asian non-Hispanic group flips between them.

People affected

Estimated range 479,019–515,308 As of 2024

Sensitivity The interval is not a confidence interval and carries no sampling error. It is a composition choice: the lower bound counts only non-Hispanic Black births, and the upper bound adds the two groups whose 2024 rates could not be published but whose most recent published ratios were the highest recorded. Non-Hispanic Asian births, 6.3 percent of the 2024 total, were deliberately left out of both bounds because the two available measures disagree about that group. The NCHS 2024 maternal mortality rate puts Asian non-Hispanic women at 18.1 against 14.2 for White non-Hispanic women, while the review-committee pregnancy-related figures for 2023 put them at 10.7 against 14.9. Including them would raise the upper bound to about 743931 on evidence that points both ways. Three further limits. The population is an annual cohort and not a cumulative one, so it does not accumulate across years. The published shares are rounded to one decimal place, which moves the lower bound by roughly plus or minus 1800 births. And the count of deaths this exposure produces is not in the interval at all: at the published rates the gap is worth roughly 147 deaths a year, derived by applying 44.8 and then 14.2 per 100000 to the 479019 births, because no source opened for this file gives a direct count of maternal deaths by race

Regional breakdown State-level maternal mortality rates are published for only 39 of 50 states, as a 2019-2023 aggregate, with counts suppressed below 9 deaths for confidentiality and below 20 deaths for data integrity, and no source opened here breaks those state rates down by race. Births by race are likewise not available by state in the sources read. Splitting the national count of Black non-Hispanic births across states in proportion to state population would be proportional allocation, and it would be especially wrong here because both births and the Black share of births are distributed very unevenly across states