NSF’s flagship fellowship program is rejecting applicants without peer review

From ScienceMag:

At least 50 students hoping to win a prestigious graduate fellowship this year from the National Science Foundation (NSF) have already been turned down without even having their research proposals reviewed. It’s an unusually high number of such rejections for a program that has long served as a steppingstone for many prominent scientists, including 50 Nobel laureates.

In recent weeks, students seeking NSF’s Graduate Research Fellowship (GRF) have received brief and identical emails from NSF saying “your application did not meet the Eligibility/Compliance requirements and is being returned without review (RWR).” An NSF spokesperson said applicants may be ineligible because they don’t meet “requirements related to degree status, field of study, degree program, or proposed research.” The spokesperson said the RWR decisions, first reported by Eos, are final and cannot be appealed.

The agency’s cryptic messaging—the spokesperson declined to answer specific questions from Science about the process—has led to speculation about what has prompted the large number of cursory rejections. A watchdog group called Grant Witness has analyzed RWR emails sent to 50 applicants and found most had proposed research in the life sciences, although the team says it has no way of knowing whether its sample is representative or how many RWR notices NSF has sent out.

“In past years, [NSF] might have excluded a handful of applications that were clearly inappropriate,” says Stony Brook University cognitive psychologist Susan Brennan, a former GRF program officer. “But this number is quite unusual. And it runs counter to the program’s goal of preparing the next generation of scientists by giving them a chance to have their ideas be subject to rigorous peer review.”

Some applicants find the rejections particularly puzzling because similar applications they submitted in 2024 did very well. A first-year graduate student who asked to remain anonymous says she was shocked to receive an RWR email late last month for essentially the same proposal she submitted in late 2024 that received an honorable mention. “It was very disappointing,” the student recalls, “since my previous proposal had received a very high score. And this one was probably even stronger, because I had tweaked it to respond to comments from the reviewers. It makes no sense.”

Another first-year graduate student whose previous proposal had also earned an honorable mention before receiving an RWR this month applied “to show my adviser that I could propose a research project based on my own ideas. It was also a way to get my feet wet and learn how the process works.”

The GRF program was one of NSF’s first initiatives after the agency was founded in 1950, supporting aspiring scientists in any field that NSF funded. But it has traveled a bumpy road since President Donald Trump took office a year ago. The 2025 class was roughly two-thirds its original target size of 2300—and it reached that number only after NSF added a second cohort of 500 students, following community outrage after NSF announced in April 2025 it would only be supporting 1000 students. The second batch of recipients skewed toward computing and quantum science, fields the Trump administration has emphasized.

The GRF was designed to support college seniors planning to attend graduate school. But over time second-year graduate students, with a year of research under their belts, became more likely to win the 3-year scholarship. Last fall, NSF announced new rules for this year’s competition that banned those more experienced students.

On social media, GRF applicants and NSF-funded researchers have speculated about the causes of the perfunctory rejections. Some wonder whether NSF judged the proposed work to be related to a particular disease, the domain of the National Institutes of Health, rather than understanding an underlying molecular mechanism. Another popular theory is that applicants used words that are anathema to the Trump administration, such as climate or diversity. Some onlookers also wondered whether NSF might be trying to reduce the number of outside reviewers needed to rate the proposals after agency program managers carry out an initial vetting of what could be up to 14,000 submissions.

The cloud of uncertainty has already had a devastating impact on students whose proposals have been summarily rejected. “Neuroscience/life sciences are still listed in this year’s solicitation [as eligible fields],” one applicant posted on Reddit in searching for an answer to why they had received an RWR email. “I am just so upset … I feel like I tried so hard and [that] none of it means anything.”

In its analysis, Grant Witness classified 12 RWRs as involving ecology, six in cell biology, five in psychology, four in neuroscience, and 14 in other subfields of biology. None of the RWRs related to work in artificial intelligence or quantum science. The rejections are evenly divided between undergraduate and first-year graduate students, says Lauren Kuehne, an independent ecologist and member of the Grant Witness team, and no single university or institution appears to have been targeted.

Whatever the total number of RWRs, there are signs that NSF may be retreating from its initial position that the rejected proposals “will not be considered further.” One student who received an RWR email tells Science that NSF rejected their first appeal but their second resulted in being told that the proposal “is being actively reviewed.”

NSF typically announces the new GRF class in April.

Read More