What is PubMed, exactly?
Adventures through the forest of US federal bureaucracy
If you’re like me, when you click through online scientific papers you’ve probably wondered what all these logos and brand names and organisations actually stand for when you read PubMed articles.
I had a vague understanding that PubMed contains free research papers, and that most US research that receives federal US funding must submit their literature for free public access on PubMed. But beyond that? Who knows? It’s just a big jumble of bureaucracy that sometimes delivers me interesting scientific and medical literature to read.
I was recently procrastinating on a more substantial blogpost which required a great deal of digging through medical research papers, and I found myself wondering about PubMed again and what it actually is. I therefore decided to write this much shorter, much lower-effort post to satiate my curiosity.
TLDR
PubMed is a database of research literature (mostly scientific papers) published in the US. Within PubMed is PubMed Central, containing millions of biomedical research papers fully accessible to the public at no charge.
The PubMed database is operated by the National Center for Biotechnology Information (NCBI), a body of the National Library of Medicine (NLM), which is an organ of the National Institutes of Health (NIH).
In other words: NIH is the agency, NLM is the division, NCBI is the department, and PubMed is the product that you actually interface with.
Cool. But what do all those other things mean, exactly?
The National Institutes of Health
In a purely capitalist system, research and development is entirely privately funded. With the exception of philanthropy, this means that in an efficient market, R&D is a proprietary effort which only receives as much funding as its returns will be commercially applicable.
Even if you are a die-hard capitalist you probably will agree that this is non-ideal; no one is going to get rich off of hepatitis B vaccinations or food safety epidemiology, but the social order really benefits from research in those fields. As such, despite the fact that the United States of America likes to imagine itself as a low-welfare small-government capitalist nation, the US federal government spends almost $50 billion1 per year on biomedical and public health research via the National Institutes of Health, or NIH.
The NIH is headquartered in Maryland (just outside of D.C.) and operates as a group of 27 federated bodies, everything from the National Institute of Arthritis and Musculoskeletal and Skin Diseases (NIAMS) to the National Institute on Alcohol Abuse and Alcoholism (NIAAA).
Some 1% of the NIH’s budget goes to a very important but relatively small institute called the National Library of Medicine (NLM).
The National Library of Medicine
The NLM is an actual literal library in Bethesda, Maryland which employs over 1700 people and has a collection of over seven million items on medicine and related sciences, according to Wikipedia. The Reading Room is open to all, so if you want to visit and read some old original 1940s text on radiation sickness or whatever, you are free to do so.
Despite this, NLM is best known for its online resources such as ClinicalTrials.gov which contains information on a quarter million human medical trials conducted worldwide and PubChem which documents millions of chemicals. NLM also runs the Radiation Emergency Management System which is the USA’s first point of contact for radiation injury/sickness information in the event of a nuclear disaster/strike.
The National Center for Biotechnology Information
Within NLM is the National Center for Biotechnology Information (NCBI) which does many things.
“Biotechnology” is one of those weird words in science that kind of means many things and isn’t very useful on its own; NCBI operates databases on human DNA sequences (GenBank), a record of nearly a billion protein sequences (Protein), a numerical taxonomy database, and more.
NCBI also operates a little side project called PubMed, which finally brings us to the end of our journey through the organised chaos of US federal bureaucracy.
PubMed
PubMed is a free online database for research papers. While tens of millions of papers have their abstracts/citations listed in PubMed, only a subset of those papers are listed in PubMed Central (PMC). Papers in PMC are free to read online and download for anyone. No subscription or institution affiliation is required.
Authors, institutions, and publications do not submit literature to PubMed directly. People at NCBI curate biomedical literature based on review criteria and decide what to index. PubMed as a database is intended to make online research easier; back in 1996 when PubMed was launched, the technology to skim through all literature across all journals was revolutionary, but researchers still needed to pay for access to the journals to read the articles.
All of that changed in 2008, when the US Congress made one of its best decisions of the 21st century and mandated that all NIH-funded research (which is most non-proprietary biomedical research in the United States) must be uploaded to a free-access database, PubMed Central, within 12 months of publication2. As of 2025 this 12 month embargo was eliminated; at the time of writing any NIH-funded paper must be available on PMC immediately.
There are still some asterisks here. Technically a lot of private US companies— pharmaceuticals and bioengineering firms in particular— do get NIH funding but are not required to submit their work to PMC. Loads of ARPA-H3 funded research cannot be accessed in PMC. The US Departments of Defense and Veterans Affairs fund their own biomedical research which sometimes isn’t required to submit to PMC either. But by and large, PMC is an incredible resource which makes the majority of US biomedical research available for free to anyone who has an internet connection.
So PMC is the subset of universally-accessible research papers in the PubMed database, which is managed by NCBI, which is a division of NLM, which is a federated institute of the NIH. Pretty interesting; even if you believe that the US federal government is a bloated mess and the federal civil service needs to be drastically reduced, you’d have to be pretty ignorant to disagree that PubMed/PubMed central is extremely socially valuable yet could only have been built through public policy. And if you think that doesn’t matter, I frankly think your personal politics are moronic and I don’t care to hear your opinion on anything else.
Remarkably, this was less than 3% of US discretionary spending in 2025. Americans sure love blowing money on missiles and tanks.
PMC was actually created in 2000, but from 2000-2008 participation in PMC was voluntary.
“Advanced Research Projects Agency for Health”. Same concept as its older Defence-based research agency, DARPA, which computer nerds are very familiar with. There’s also ARPA-I for infrastructure research.
<3 PubMed