Mathematics is the language of science and the foundation of quantitative reasoning across nearly every academic discipline. It is also one of the most versatile undergraduate majors available, producing graduates who become research mathematicians, Wall Street quantitative analysts, software engineers, statisticians, data scientists, actuaries, economists, and educators. The career pathways from a mathematics degree are more diverse than those from almost any other undergraduate major, which is both the discipline’s great strength and the central challenge in interpreting earnings data for mathematics programs.
Approximately 25,000 mathematics bachelor’s degrees are awarded annually in the United States, across more than 518 programs with sufficient scale and data to be included in this ranking. That total includes degrees in pure mathematics (CIP 27.01), applied mathematics (CIP 27.03), statistics (CIP 27.05), and related mathematical sciences, all fields housed within the same academic departments at most institutions.
How We Built the Ranking
The approach evaluates all 518 institutions across five components. The most important methodological note concerns the Earnings & ROI component, which applies a two-tier design for mathematics.
| Component | Weight | Data Source |
| Major Emphasis | 12% | IPEDS Degree Completions (all CIP 27.xx) |
| Program Scale | 13% | IPEDS Degree Completions (log) |
| Academic Rating | 30% | IPEDS / Common Data Set |
| Earnings & ROI (Tier 1) | 25% | Scorecard field earnings 55% + inst. 15% + ROI 20yr 15% + ROI 40yr 15% |
| Earnings & ROI (Tier 2†) | 25% | Inst. earnings 55% + ROI 20yr 22.5% + ROI 40yr 22.5% (where field earnings unavailable) |
| PhD Productivity | 20% | NSF SED: “Mathematics,” “Applied mathematics,” and “Mathematics and statistics, other” summed |
The two-tier C4 design. Field-specific mathematics earnings are available for 61 percent of the 518 programs in this universe. For the 39 percent of programs without available field earnings, including Princeton, Yale, Caltech, Harvey Mudd, and Penn, where most graduates proceed directly to doctoral programs and the direct-employment cohort is too small for Scorecard disclosure, we apply the Tier 2 formula: institutional 10-year earnings receive a 55 percent weight, with the remaining 45 percent distributed between the two ROI measures. This is more appropriate than imputing missing values at the universe median, which would penalize high-quality programs for having small direct-employment cohorts. In the Top 25 table below, programs using the Tier 2 formula are marked with a dagger.
PhD Productivity (20 percent) sums doctoral alumni across three NSF fields: “Mathematics,” “Applied mathematics,” and “Mathematics and statistics, other.” The Tier 2 treatment matters most for the PhD Productivity leaders. Programs like Princeton, Caltech, and Harvey Mudd that send very large shares of graduates to doctoral programs naturally have suppressed field earnings, and the two-tier formula ensures they receive appropriate credit for their economic outcomes.
The Top 25
The top 25 programs are shown below. Programs marked with a dagger use the Tier 2 C4 formula, meaning field-specific earnings were unavailable and the component instead reflects institutional earnings and ROI. That treatment places Yale (#7), Princeton (#9), Caltech (#15), and Harvey Mudd (#27) at positions consistent with their academic quality and doctoral pipelines, rather than at the artificially low positions that median imputation would have produced.
| Rank | Institution | State | Type | C4 | Score |
| 1 | Harvard University | MA | Private | T1 | 96.8 |
| 2 | University of Chicago | IL | Private | T1 | 95.3 |
| 3 | Massachusetts Institute of Technology | MA | Tech | T1 | 95.2 |
| 4 | Brown University | RI | Private | T1 | 95.0 |
| 5 | Carnegie Mellon University | PA | Private | T1 | 95.0 |
| 6 | Duke University | NC | Private | T1 | 93.3 |
| 7 | Yale University | CT | Private | T2† | 93.0 |
| 8 | Stanford University | CA | Private | T1 | 92.1 |
| 9 | Princeton University | NJ | Private | T2† | 91.6 |
| 10 | Rice University | TX | Private | T1 | 91.5 |
| 11 | Dartmouth College | NH | Private | T1 | 91.5 |
| 12 | Columbia University | NY | Private | T1 | 91.3 |
| 13 | University of Notre Dame | IN | Private | T1 | 90.6 |
| 14 | Northwestern University | IL | Private | T1 | 90.5 |
| 15 | California Institute of Technology | CA | Tech | T2† | 89.8 |
| 16 | Swarthmore College | PA | Liberal Arts | T1 | 89.7 |
| 17 | Johns Hopkins University | MD | Private | T1 | 89.2 |
| 18 | Williams College | MA | Liberal Arts | T1 | 89.2 |
| 19 | University of Pennsylvania | PA | Private | T2† | 88.5 |
| 20 | Amherst College | MA | Liberal Arts | T1 | 88.5 |
| 21 | Cornell University | NY | Private | T1 | 88.1 |
| 22 | University of California-Berkeley | CA | Public | T1 | 87.6 |
| 23 | New York University | NY | Private | T1 | 87.5 |
| 24 | University of California-Los Angeles | CA | Public | T1 | 87.2 |
| 25 | University of Michigan-Ann Arbor | MI | Public | T1 | 87.0 |
Table 1. Top 25 Undergraduate Mathematics Programs, 2026 College Transitions Ranking. † = Tier 2 C4: field-specific earnings unavailable; formula uses institutional earnings + ROI. LAC = Liberal Arts College.
Harvard University (#1, 96.8) leads the mathematics ranking, combining near-perfect performance across all five components: 99.8 on Major Emphasis, 96.5 on Program Scale, 93.1 on Academic Rating, 98.6 on Earnings & ROI (Tier 1, field earnings of $166,324 reflecting the finance and technology premium), and 98.5 on PhD Productivity (95 math/stats doctoral recipients at 13.4 per capita). Harvard’s mathematics department is one of the world’s foremost research centers, consistently producing graduates who lead careers in research mathematics, theoretical computer science, and quantitative finance.
University of Chicago (#2, 95.3) leads all programs on PhD Productivity (99.2, reflecting 136 math/stats doctoral recipients, the highest raw count in the dataset) and benefits from Chicago’s mathematics-for-its-own-sake academic culture that consistently produces graduates unusually well-prepared for advanced doctoral work.
MIT (#3, 95.2) and Brown (#4, 95.0) complete the top tier, each scoring above 96 on PhD Productivity. Carnegie Mellon (#5, 95.0) achieves a perfect 100 on Major Emphasis, the highest in the full dataset, reflecting the concentration of CMU’s undergraduate enrollment in mathematical sciences, aided by field earnings of $142,883 (the finance and quantitative technology premium for CMU mathematics graduates).
Yale (#7, 93.0) and Princeton (#9, 91.6) are the clearest illustrations of what the two-tier treatment does. Yale’s field earnings are unavailable in the Scorecard; its Tier 2 C4 score of 96.7 reflects strong institutional earnings and ROI, placing it at #7. Princeton’s field earnings are suppressed by a very small direct-employment cohort, reflecting that Princeton mathematics graduates proceed to doctoral programs at one of the highest rates of any program in the country. Its Tier 2 C4 score of 99.1 is supported by a 94.5 Academic Rating (second-highest in the dataset) and 99.1 on PhD Productivity (126 doctoral recipients at 22.2 per capita). Princeton’s composite of 91.6 reflects a program among the strongest in the country.
What Separates the Best Programs?
The Liberal Arts College Phenomenon
One of the most distinctive features of the mathematics ranking is the strong presence of small liberal arts colleges in the top 25. Swarthmore (#16), Williams (#18), and Amherst (#20) all rank in the top quartile, above most large public research universities. Liberal arts colleges achieve this through three structural advantages: very high Major Emphasis scores (mathematics accounts for a larger fraction of total degrees at a small college than at a comprehensive research university), strong academic environments, and high per-capita PhD productivity. Swarthmore sends 22.5 mathematics doctoral recipients per 1,000 undergraduates over the decade (fourth-highest per-capita rate in the dataset). Williams sends 17.4, and Amherst sends 9.2.
Caltech and Harvey Mudd: The Per-Capita Leaders
California Institute of Technology (#15, 89.8) and Harvey Mudd College (#27, 86.6) are the clearest illustrations of why the two-tier approach matters for mathematics. Both have suppressed field earnings, so their C4 is based on institutional earnings and ROI, which captures the economic value of their degrees without penalizing them for sending nearly all graduates to doctoral study. Caltech’s 81 math/stats doctoral recipients at a per-capita rate of 82.1 is the highest rate of any program in the country. Harvey Mudd’s 56 recipients at 60.8 per capita is second. Both programs produce doctoral mathematicians at rates roughly five times higher than the next comparable programs, and the two-tier C4 ensures their composite scores reflect that.
The Large Public Research Programs
UC Berkeley (#22, 87.6) leads all public universities, with the highest raw PhD count in the dataset (171) and a near-perfect 99.0 on Program Scale (1,013 math degrees over three years). UCLA (#24, 87.2) leads all programs on Program Scale (a perfect 100, reflecting 2,177 math degrees over three years). Michigan (#25, 87.0) rounds out the top 25 with 97.5 on Program Scale and 90.0 on PhD Productivity (77 doctoral recipients).
Patterns, Themes, and What They Mean for Your Students
Mathematics earnings require more careful interpretation than any other field in this ranking series. The Tier 1 field earnings figures at elite programs (Harvard $166,000, MIT $181,000, CMU $143,000) reflect a small cohort of students who chose finance and technology careers over graduate school, not a general statement about mathematics graduate earnings. The Tier 2 programs (Princeton, Yale, Caltech, Harvey Mudd, Penn) have suppressed field earnings precisely because their graduates go to graduate school at high rates. The two-tier approach ensures neither group is penalized for the structure of its graduates’ career choices.
The choice between a liberal arts college and a research university is consequential for mathematically talented students. Students planning doctoral study in mathematics are often better served by a small liberal arts college with a faculty deeply committed to undergraduate mentorship than by a large research university where faculty attention is directed primarily toward graduate students. Swarthmore, Williams, Amherst, Pomona, and Harvey Mudd produce per-capita doctoral mathematicians at rates that rival or exceed MIT and Harvard. Students seeking broader career options in finance, technology, or data science are often better served by large research universities with strong industry recruiting.
The Putnam Competition is a signal of program quality that rankings cannot fully capture. Programs that consistently place students among the top finishers in the William Lowell Putnam Mathematical Competition, among them MIT, Harvard, Stanford, Caltech, Carnegie Mellon, Princeton, and Chicago, have built the instructional culture and peer environment that makes serious mathematical work at the undergraduate level possible. Counselors working with highly talented mathematics students should include Putnam performance in their program evaluation.
Applied mathematics and statistics are increasingly important for career outcomes. Programs with strong applied mathematics and statistics tracks, including MIT, CMU, NYU (the Courant Institute), Brown, Berkeley, and Chicago, prepare students for the data science, machine learning, and quantitative finance careers that absorb the majority of non-academic mathematics graduates. Understanding whether a program is oriented primarily toward pure or applied mathematics matters more for career outcomes than composite ranking position.
Mathematics remains the foundational discipline for a wide range of intellectual pursuits and careers. The programs that lead this ranking have built the mathematical research cultures, faculty depth, and student communities that make substantive mathematical education possible at the undergraduate level. Students who choose carefully, matching a program’s orientation and culture to their own goals, will find a field with strong prospects across both academic and industry paths.
