Then, in 2025, the freezes and terminations began. Each dot is one grant that was frozen or terminated. The count does not stop at 6 terminations. It keeps going.
More than grants 5,362 were frozen.
734 were terminated In total, 6,096 NIH grants were impacted.Before we dive into that, we need some background.
The National Institutes of Health is the largest public funder of biomedical and behavioral research in the world. Its work doesn't stay in the lab. It becomes the treatments people depend on.
As part of the U.S. Department of Health and Human Services, the NIH invests tens of billions of dollars annually to support research at universities, hospitals, and research institutions across the country. Its history traces back to 1887 (NIH, 2025a).
Its reach goes well beyond any single breakthrough. Tap any area below to see what the research it funds delivers. Except where noted, every figure comes from the NIH's own public accounting.
The NIH puts a dollar value on that progress: each 1% reduction in cancer deaths carries a present value of nearly $500 billion to current and future generations of Americans.
NIH Cancer Research Click to Open in New TabBroader evidence-based addiction programs can return more than $58 for every $1 spent, thanks in part to NIH-supported work connecting the science of treatment to real-world practice. In one example, NIH-supported research showed that giving medications for opioid use disorder to people in prison lowers overdose risk after release, and helped drive policy changes such as Medicaid coverage in the 90 days before release.
Addiction Treatment Click to Open in New Tab Cost Benefits of Substance Abuse Prevention, Click to Open in New TabFor children born in 2009, the CDC estimates childhood vaccination will cut direct health care costs by $13.5 billion and save $68.8 billion in indirect costs over their lifetimes. NIH supported the Hib vaccine alone, estimated to save more than $1.8 billion for that single birth year, and early cochlear implants developed in part from NIH-supported research can save more than $30,000 per child.
Economic Review of Childhood Vaccination, Click Open in New Tab Cochlear Impants in Children, Click to Open in New TabLead early in life irreversibly disrupts brain development. The decline was achieved through control of lead sources, emission reductions, federal regulation, and applied public health work. NIH-funded environmental research has also produced remediation technologies estimated to save over $100 million in cleanup costs.
NIH Lead Exposure Research Click to Open in New TabNIH-funded research identified adolescence as a period of increased vulnerability to the harms of drug use, and a key window for prevention. If implemented nationwide, school-based substance use prevention programs are estimated to save $18 per $1 invested, with 1.5 million fewer youth starting substance use and, on average, a two-year delay for those who do.
Cost Benefits of Substance Abuse Prevention, Click to Open in New Tab Research on Drug Use Prevention in Children, Click Open in New Tab Research on Adolescent Drug Use Risks, Click to Open in New TabSource: National Institutes of Health. Figures reflect NIH's own published estimates of the impact of research it funds.
A grant is public money NIH gives an institution to answer one scientific question over several years. The money arrives one year at a time, not all at once. That single fact is what makes an interrupted grant so costly.
Once a year, Congress passes a law setting how much money NIH receives. Under the Constitution, only Congress can decide federal spending.
Once that law is signed, the money is supposed to be spent as written.
The Treasury holds the nation's money and pays it out as agencies commit it. It is the account, not the decision-maker.
NIH divides its funding among 27 institutes and centers, each focused on a field of health. Most of that money does not stay inside NIH.
NIH sends most of its money to institutions such as universities, medical schools, and hospitals. Each grant names a lead scientist, the principal investigator, but the institution holds the funds, hires the staff, and runs the project.
This is why an interrupted grant strands a whole team, not one person.
In early 2025 the Trump administration paused and withheld NIH funding, releasing roughly $8 billion less than in prior years.
In August 2025 the Government Accountability Office, Congress's nonpartisan watchdog, found this broke the Impoundment Control Act of 1974, the law that says a president cannot refuse to spend money Congress has approved. The Trump administration disputes this, and the case is still in court.
The team does the work, files a progress report, and receives the next year's installment. It stays here until it finishes, is frozen, or is terminated.'
Fewer than 6 NIH grants were terminanted from 2012 until 2025.
And how much was already invested in the projects that were completely terminated?
Every dollar NIH commits to a grant ends up in one of three places once the grant is cut:
$926 Million was committed to these grants. $620 million of it had already been paid out for work that then stopped and that money is gone regardless of what happens next. The government formally reclaimed $151 million, about 14 cents of every dollar it had committed.
These are the 734 grants with spending on record. The 515 without it skew older and would have spent more, so $926 Million is a floor, not a ceiling: the real total is higher, not lower.
Whichever region or institution a grant belonged to, a disruption carried the same underlying cost: work already paid for, and work that would now never be finished.
Every disrupted grant sits somewhere on a spectrum. Some were frozen only briefly before funding resumed, others were cut off for good. But even a grant that is eventually restored can leave real waste behind: salaries already paid, experiments half-run, and labor lost. The amounts calculated cover the roughly 800 disrupted grants with complete spending records, and should be read as a floor rather than a full accounting, because the grants missing spending data tend to be the oldest and most established.
The money is concentrated in grants that had been running for years. A first-year grant that gets cut costs the public a few tens of thousands of dollars; a fifth-year grant costs well over a million. Whichever group you highlight, the shape holds: the cuts landed on mature work.
(More than halfway throughtheir scheduled duration)
A grant has a scheduled start and end. The longest running grant that was terminated by the Trump administration was 44 years long.
Half the grants that were cut were more than 59% of the way through their scheduled work. These weren't proposals being declined. They were projects with staff hired, patients enrolled, and years of work behind them.
Most of the grants frozen by the Trump administration were slowly restored.
But some were not.
The graph above shows the number of grants frozen and restored over time.
Beginning in February 2025, the Trump administration began terminating active NIH research grants, cutting off researchers mid-project. Over the following months, more than 6,000 grants were disrupted. Some were restored after courts intervened. Others were not. After courts ruled the terminations illegal, funding levels slowly recovered
Many reinstated (restored) grants have come substantially through specific court cases, and litigation eligibility tends to cluster by which states attorneys general or named institutions joined those suits.
The map on the left shows the count distribution of terminations. The map on the right shows the rate distribution of terminations. Rates are calculated a state/region's basis for detail.
Gross termination shows that alomst every state got hit, but Still-Terminated and Reinstated, tells a much different story.
Several public health-related research areas rank among the NIH research areas with the highest grant termination rates.
These trends are even more apparent when comparedside-by-side
Among grants flagged in this dataset, public health related research areas consistently saw the largest share end in disruption rather than restoration. The size of a research area was not strongly associated with that share, so the pattern is not simply an artifact of big fields having more grants. Instead, specific topics stand out: health disparities, minority health, rural health, infectious diseases, and vaccine related research were disproportionately affected.
This interactive graphic shows the number of grants disrupted in the different research areas. There are fluctuations in disruptions as courts restored funding to the grants.
This chart shows that private sector research institutions had their frozen grant money returned to them much more quickly than their public counterparts. Similarly, non-minority serving research institutions had their frozen grant money returned to them much more quickly than minority-serving institutions.
This analysis summarizes key patterns in NIH grant disruptions:
Beginning in early 2025, thousands of active NIH grants were abruptly frozen or terminated, cutting off research already underway. The courts later forced much of that funding to be restored, but restoration was not the same as prevention: by the time money flowed again, the disruption had already done its damage.
That damage did not land evenly. Public health research, minority-serving institutions, and grants containing certain flagged language were disrupted at higher rates and, once frozen, stayed frozen longer. Whatever the intent behind the cuts, the burden fell hardest on a specific and identifiable set of the affected grants rather than being spread across the research enterprise as a whole.
And the disruptions were expensive in a way a reversal cannot fix. Grants were not cut at their start, before resources had been committed; terminated grants had typically passed the 57% mark of their project period. The government had already paid for most of the work and stood to lose the payoff of finishing it. Cutting a grant this late is the difference between canceling a house before construction and abandoning it with the roof half on: over half a billion dollars had already been spent on work that no termination can un-spend.
Even that figure understates the loss. It captures money, but not the years of training interrupted, the patient studies halted mid-course, or the research questions left unanswered. And it is a conservative floor, drawn only from grants with complete records, with the oldest and most established grants missing entirely. The full cost of stopping science midway is larger than any single number in this piece can show.
