Fulbright publishes the application count behind every institution’s grant total. Almost nobody looks at that column.
On February 3, 2026, the Fulbright Program named 114 colleges and universities as Top Producing Institutions for the 2025-2026 award year. Announcements followed within days. Brown reported 41 student grants and a tenth straight year in the top three. Princeton reported 25. Michigan reported 25 student awards alongside 2 faculty Scholar awards, and noted it has been a top producer for more than fifteen years.
Every one of those announcements reported the same statistic: how many students won. Not one reported how many applied.
The Fulbright Program publishes both. Its Top Producing Institutions data table carries a Grants column and an Applications column, side by side, for every institution that had a student selected. The second column changes the meaning of the first, and it goes almost entirely unread.
A note on timing before the numbers. The 2025-2026 table is the most recent nationally comparable dataset and it is the one used throughout this article. Individual colleges have already begun announcing their 2026-2027 results, and Brown reported 39 awards in June 2026, but the State Department does not release the national table for that cycle until 2027. Campus announcements and the national dataset run about eight months apart.
What the Published List Actually Measures
The 99 baccalaureate institutions appearing in Fulbright’s published 2025-2026 table account for 403 grants offered against 1,925 applications submitted. Pooled across all of them, 20.9 percent of applications produced a grant. The median college in the table sent 14 applications and won 2. Colleges whose students applied and won nothing do not appear, so this is the population of schools with at least one recipient rather than every school that tried.
Thirty-six of those 99 carry the Top Producer designation. The designation is a count of awards offered, ranked within Carnegie classification. A college earns it by winning a lot of grants, which it can do either by winning at a high rate or by sending a very large number of students and winning at an ordinary one. The published list does not distinguish between those two paths, and a family reading it cannot tell which one a given school took.
Pitzer College leads the baccalaureate category with 18 grants. It sent 76 applications to get them, a hit rate of 23.7 percent. Bowdoin College is tied for second with 17 grants, from 46 applications, a hit rate of 37.0 percent. Both colleges appear on the same list, in nearly the same position, describing very different experiences for the student who applies.
Applications and grants for all 99 baccalaureate institutions. Source: Fulbright U.S. Student Program, Top Producing Institutions data table, 2025-2026 award year, released February 3, 2026.
Twenty-Nine Applications, One Grant
The most striking contrast in the table sits inside a single consortium. Four Claremont colleges appear in the 2025-2026 baccalaureate data, all in Southern California, all sharing a campus footprint, all drawing from the same regional and national applicant pool.
Pitzer sent 76 applications and won 18. Pomona sent 75 and won 17, a rate of 22.7 percent. Scripps sent 28 and won 7, a rate of 25.0 percent. Claremont McKenna sent 29 and won 1.
Take that carefully, because the arithmetic is less decisive than the numbers look. A test of homogeneity across all four colleges returns p = 0.10, so the four as a group cannot be shown to differ beyond chance in a single cycle. Pitzer, Pomona, and Scripps are statistically indistinguishable from one another. Claremont McKenna against each of its three neighbors lands near p = 0.02 individually, which is suggestive and not conclusive once you account for how many comparisons a table this size permits.
What the consortium does establish is that a gap this size can open between colleges whose students are drawn from the same pool. Differences in overall institutional selectivity or student-body academic strength are unlikely to explain a gap this large on their own. Several other things could contribute, including which countries and grant types students pursue, the mix of research proposals against English teaching assistantships, applicant self-selection, and how each campus supports its applicants. The published table shows none of them.
The pattern appears outside Claremont as well. Reed College and Hamilton College each submitted 43 applications and each won 5. Bowdoin submitted 46, three more than either, and won 17. Swarthmore submitted 31 and won 4, a rate of 12.9 percent, while Amherst submitted 52 and won 17, a rate of 32.7 percent. These are peer institutions by any selectivity measure a family would use.
One Award Year Cannot Rank These Colleges
Among the 36 colleges that submitted at least 20 applications, observed hit rates run from Bowdoin’s 37.0 percent down to Claremont McKenna’s 3.4 percent, a spread of nearly eleven to one. On a chart that looks decisive. It is not.
A chi-square test of homogeneity across all 36 returns p = 0.067. That falls short of the conventional threshold, which means a single award cycle does not contain enough evidence to establish differences in the colleges’ underlying rates across the group. That is not the same as showing the rates are equal. It means the ordering visible in the chart below cannot be read as a ranking of underlying performance.
A few comparisons at the extremes do hold. Fisher’s exact test puts Bowdoin against Claremont McKenna at p = 0.0007 and Amherst against Claremont McKenna at p = 0.0019. Both survive a Bonferroni correction for the eleven comparisons named in this article. Bowdoin against Reed returns p = 0.007, nominally significant and short of that corrected threshold. Bowdoin against Pitzer returns p = 0.15, which is to say the two are not distinguishable, despite a fourteen-point gap in their observed rates.
So this is not a better ranking than the Top Producer list. It is a demonstration that one year of fellowship data cannot rank colleges at all, and that the badge is measuring something else entirely.
Counselors should read the chart below as a picture of uncertainty. A college three places higher than another is not measurably better at this, and in most cases neither is a college fifteen places higher.
Hit rate with 95 percent Wilson confidence intervals, institutions with at least 20 applications. A homogeneity test across all 36 returns p = 0.067. Source: Fulbright U.S. Student Program, Top Producing Institutions data table, 2025-2026 award year.
Higher Volume Does Not Predict a Lower Hit Rate
One reasonable theory would predict that the high-volume colleges post lower rates, because pushing more students through the process means pushing weaker ones. Among the 58 baccalaureate colleges with at least 10 applications, the correlation between application volume and hit rate is 0.146 by Pearson and 0.220 by Spearman. There is a faint positive tilt and nothing that would let you predict one from the other.
Read that as a statement about this cross-section and not about cause. Volume and conversion sit on close to independent axes across these colleges. It does not follow that a particular college could double its applications and hold its rate.
Oberlin sent 76 applications and converted 21.1 percent, almost exactly the pooled figure. Pomona sent 75 and converted 22.7 percent. Neither shows an obvious rate penalty associated with its volume. Meanwhile Claremont McKenna’s 29 applications and Colby’s 20 are modest volumes attached to the lowest observed rates in the set. A low hit rate in this data is not explained by a college being unusually permissive about who applies.
A Third Measure Reorders the List Again
Grants per 1,000 undergraduates measures Fulbright production against institutional size. It is not a student’s odds and not a per-senior rate, since the Fulbright U.S. Student Program draws from graduating seniors, current graduate students, and recent alumni while the denominator here is undergraduate enrollment. As a size adjustment it does useful work, and it produces a third ordering.
Pitzer, with 1,242 undergraduates, records 14.49 grants per 1,000, the highest figure in the category by a wide margin. Pomona follows at 10.00, Bowdoin at 9.04, Amherst at 8.88, Williams at 8.04. Wabash College, with 866 undergraduates and 6 grants from 12 applications, records 6.93 per 1,000 and the highest observed hit rate in the table at 50.0 percent, on a volume far too small to carry weight.
The three measures disagree by a lot for individual colleges. Occidental College ranks 14th of 14 on grants won, 4th on hit rate, and 9th on grants per 1,000, a ten-place spread. Smith College spans nine places, Wesleyan eight, Pitzer and Pomona seven each. Williams is the only college among the top 14 that lands in the same position on all three.
That disagreement is the practical point. Three defensible ways of reading the same table produce three different lists, and the published list uses only one of them.
Rank position on each of three measures, top 14 colleges by grants won. Source: Fulbright U.S. Student Program, 2025-2026 award year; undergraduate enrollment from IPEDS, fall 2024.
What Campus Advising Does to These Numbers
Applying to Fulbright through an institution is a multi-stage campus process involving nomination. Brown, for example, publishes a sequence running from an interest form in June through a draft of the statement of grant purpose, pairing with a fellowship advisor, and campus endorsement ahead of the national October deadline. Other campuses run lighter processes. The published data records none of this.
That gap limits what any hit rate can be read to mean. A high conversion rate could reflect an unusually strong applicant pool, unusually effective advising, a favorable mix of host countries and grant types, or a combination. The public table separates none of those, and it does not report how much support any individual applicant received before submitting.
So the application count tells a family considerably more than the Top Producer badge does, and still stops well short of explaining why one college’s students convert at three times the rate of a peer’s. Anyone who wants that answer has to ask the college, which is easier than it sounds.
The Top Producer Badge Is a Count, Not a Rate
The badge and the application threshold pick out different groups of colleges, and by coincidence both groups number 36. Thirty-six of the 99 institutions carry the Top Producer designation. A separate 36 submitted at least 20 applications. Twenty-nine colleges appear in both groups.
Among those 36 higher-volume colleges, 29 hold the badge and 7 do not, and the badge does not track the hit rate. Hamilton and Reed both convert at 11.6 percent and both hold it. Colorado College converts at 10.7 percent and does not, on one fewer grant. Swarthmore converts at 12.9 percent and holds it; Dickinson converts at 10.7 percent and does not.
The full table for the 36 higher-volume colleges is below, sorted by hit rate. It is the version a counselor can check a specific school against, with the caveat from the previous section attached: the ordering within it is not reliable.
| College | Grants | Applications | Hit rate | Grants / 1,000 UG | Top Producer |
| Bowdoin College | 17 | 46 | 37.0% | 9.04 | Yes |
| University of Portland | 8 | 22 | 36.4% | 2.68 | Yes |
| Amherst College | 17 | 52 | 32.7% | 8.88 | Yes |
| Smith College | 10 | 32 | 31.2% | 3.92 | Yes |
| Occidental College | 9 | 30 | 30.0% | 4.78 | Yes |
| Williams College | 17 | 62 | 27.4% | 8.04 | Yes |
| Wesleyan University | 11 | 42 | 26.2% | 2.89 | Yes |
| Middlebury College | 7 | 27 | 25.9% | 2.54 | Yes |
| Scripps College | 7 | 28 | 25.0% | 6.27 | Yes |
| Wellesley College | 11 | 45 | 24.4% | 4.57 | Yes |
| Macalester College | 7 | 29 | 24.1% | 3.27 | Yes |
| Pitzer College | 18 | 76 | 23.7% | 14.49 | Yes |
| Kenyon College | 8 | 34 | 23.5% | 3.56 | Yes |
| Pomona College | 17 | 75 | 22.7% | 10.00 | Yes |
| College of the Holy Cross | 12 | 53 | 22.6% | 3.84 | Yes |
| Bryn Mawr College | 5 | 23 | 21.7% | 3.68 | Yes |
| Oberlin College | 16 | 76 | 21.1% | 5.52 | Yes |
| Carleton College | 14 | 68 | 20.6% | 6.58 | Yes |
| Davidson College | 8 | 40 | 20.0% | 4.28 | Yes |
| Vassar College | 4 | 21 | 19.0% | 1.62 | Yes |
| Haverford College | 5 | 29 | 17.2% | 3.49 | Yes |
| Barnard College | 11 | 65 | 16.9% | 3.15 | Yes |
| Bates College | 9 | 54 | 16.7% | 5.11 | Yes |
| Colgate University | 8 | 52 | 15.4% | 2.51 | Yes |
| Grinnell College | 4 | 26 | 15.4% | 2.24 | Yes |
| Swarthmore College | 4 | 31 | 12.9% | 2.46 | Yes |
| Bard College | 4 | 31 | 12.9% | 1.64 | Yes |
| Hamilton College | 5 | 43 | 11.6% | 2.46 | Yes |
| Reed College | 5 | 43 | 11.6% | 3.71 | Yes |
| Colorado College | 3 | 28 | 10.7% | 1.48 | No |
| Dickinson College | 3 | 28 | 10.7% | 1.30 | No |
| Colby College | 2 | 20 | 10.0% | 0.83 | No |
| Denison University | 2 | 21 | 9.5% | 0.84 | No |
| Trinity College (CT) | 2 | 21 | 9.5% | 0.91 | No |
| United States Military Academy | 2 | 22 | 9.1% | 0.45 | No |
| Claremont McKenna College | 1 | 29 | 3.4% | 0.72 | No |
Baccalaureate institutions with at least 20 Fulbright U.S. Student applications, 2025-2026 award year. Ordering within the table is not statistically reliable; see the accompanying section. Undergraduate enrollment from IPEDS, fall 2024.
What This Data Cannot Tell You
The table covers baccalaureate institutions only. The Fulbright Program publishes the same two columns for research universities, master’s institutions, and special-focus schools, but those views load through the site’s interactive filters and are not reachable in a static page pull, so Georgetown, Brown, Harvard, and every other large university sit outside this analysis. Nothing here should be read as a comparison between liberal arts colleges and research universities.
These are awards offered, not accepted. Brown reported 41 offers and 37 acceptances for the same cycle, so acceptance runs high but not universally, and the grant counts here slightly overstate the number of students who went.
Anyone extending this across years has a definitional break to handle. Fulbright’s current table reports awards offered for recent cycles, while figures prior to 2021-2022 reflect awards accepted. The two are not directly comparable.
A table of 36 colleges permits 630 pairwise comparisons. Any reader who goes looking for a difference will find several that clear p = 0.05 by chance alone. The tests reported above cover only the comparisons this article names.
This is also one award year, which is the limitation that drives all the others. Fellowship offices change staff, and a cycle at a small college turns on two or three students. Several years of application counts would settle most of what this analysis has to leave open, and Fulbright does not publish them in an accessible form.
What to Ask, and When
For a counselor sitting with a family that cares about postgraduate fellowships, the useful question is not whether a college appears on the Top Producer list. Thirty-six baccalaureate colleges do, and the designation records award volume rather than a student’s prospects.
Ask the college three things. How many students applied for Fulbright last year, and how many won. What the campus process looks like between spring of junior year and the October national deadline, and who runs it. Whether a full-time staff member handles fellowship advising, or whether it sits on top of a faculty member’s teaching load.
A college that answers all three without hesitating is telling you something. A college that answers only the first, in the form of the number it put in a press release, is telling you something too.
Families can raise this at any information session, and the answer arrives faster than a data table. The grant count tells you how many students won. The application count tells you what that number means. It is the number almost nobody volunteers.
Fulbright grant and application counts: Fulbright U.S. Student Program, Top Producing Institutions data table, 2025-2026 award year, released February 3, 2026, retrieved September 10, 2026. Figures reflect awards offered rather than accepted; Fulbright notes that data prior to 2021-2022 reflect awards accepted. Analysis covers the 99 baccalaureate institutions listed in that table and excludes the doctoral, master’s, and special-focus categories, which are served through interactive filters that were not retrievable. Undergraduate enrollment: IPEDS Institutional Panel, fall 2024. Confidence intervals shown in the second chart are Wilson score intervals at 95 percent; pairwise comparisons in the text use Fisher’s exact test, and the homogeneity tests are chi-square tests of independence. Correlation figures use the 58 baccalaureate institutions with at least 10 applications. Institutional announcement figures are drawn from Fulbright Program, Brown University, Princeton University, and University of Michigan announcements dated February 3 and 4, 2026, and from Brown University, June 24, 2026. Campus process description from Brown University Fellowships, Fulbright U.S. Student Program page, retrieved September 10, 2026.
