← Back to blog
11 min read
13 U.S. Colleges Where the Median Grad Walks Into a $100K+ Job (2026 ROI Data)
June 26, 2026 · ResuAI Editorial
If you only look at the brand names, U.S. college rankings feel rigged. Harvard, Yale, Stanford, MIT — same handful of schools dominate every "best of" list a high-schooler sees.
The salary data tells a different story. The college that produced the highest median starting salary in the United States in the most recent reporting year wasn't Harvard or Yale. It was a STEM-only liberal arts school in Claremont, California with fewer than 1,000 students total, and the class of 2024 walked into a $117,500 median starting salary (Town & Country 2025 ROI ranking, citing Harvey Mudd's own senior outcomes report).
That same year, the University of Pennsylvania — Ivy League, 10,300 undergrads, $91,112 sticker price — posted a class-of-2024 median starting salary of $100,200 (same source, Penn senior outcomes).
A school with 11x fewer undergrads is out-earning the Ivy by $17,000 a year, out of the gate.
This isn't a fluke. It's a pattern. Below is the 2026 map of the 13 U.S. colleges where graduates command the highest median salaries — anchored to PayScale's College Salary Report, Georgetown CEW's ROI 2025 dataset, the Wall Street Journal's 2026 college ROI ranking, and each school's own published class-of-2024 outcomes.
We'll also unpack the part nobody tells high-schoolers: the school you go to is a 30-second filter on your resume, but everything after that filter still runs on the resume itself. More on that in the bridge section.
The 13 schools that mint the highest-earning graduates (2026)
Here's the full picture in one table. Five columns we care about: class size, who they let in, what their grads earn day one, what they earn mid-career, and what the degree is actually worth in 20 years of compounded earnings (Georgetown CEW's net present value figure, the most rigorous ROI number publicly available).
| College | Class Size (UG) | Acceptance Rate | Median Starting Salary (Class of 2024) | Mid-Career Median (PayScale) | 20-Year NPV (Georgetown CEW) |
|---|---|---|---|---|---|
| MIT | ~4,600 | 4.52% | ~$110,200 (PayScale) | $196,900 | $1,881,000 |
| Harvey Mudd College | ~900 | 12.66% | $117,500 | $185,900 | $1,819,000 |
| Caltech | ~970 | 3.14% | $110,000–$119,000 | ~$155,000+ | $1,719,000 |
| Olin College of Engineering | ~400 | 21.66% | ~$85,000+ | n/a (89% placed) | $1,709,000 |
| Princeton University | ~5,800 | 4.62% | ~$95,000+ | $194,100 | $1,635,000 |
| Stanford University | ~7,900 | 3.61% | ~$100,000+ | ~$170,000+ | $1,630,000 |
| University of Pennsylvania | ~10,300 | 5.40% | $100,200 | ~$155,000+ | $1,557,000 |
| Carnegie Mellon University | ~7,800 | 11.66% | ~$110,000+ (CS/SCS) | ~$160,000+ | $1,527,000 |
| Babson College | ~2,800 | 17.00% | $77,681 (avg) | $181,400 | $1,523,000 |
| Bentley University | ~4,200 | 45.05% | $72,000 | ~$130,000+ | $1,485,000 |
| Georgia Institute of Technology | ~19,200 | 12.74% | ~$85,000+ (College of Engineering) | ~$130,000+ | $1,417,000 |
| Claremont McKenna College | ~1,380 | 9.59% | ~$80,000 | ~$155,000+ | $1,407,000 |
| Stevens Institute of Technology | ~4,000 | 47.58% | $84,800 (avg) | ~$130,000+ | $1,404,000 |
A few notes on the data so you can argue with the right person if you disagree:
- Class size is undergraduate enrollment only. Some of these schools are heavily graduate-skewed (MIT, Stanford, Penn) — the salary numbers below all refer to bachelor's-only outcomes per PayScale's methodology.
- Acceptance rate is Class of 2030 admit data where reported (College Kickstart Class of 2030 tracker), otherwise the most recent published rate.
- Median starting salary is each school's published class-of-2024 senior outcomes survey where available; otherwise PayScale's "early career" figure (0-5 years of experience with a bachelor's as the highest degree).
- 20-year NPV is from Georgetown CEW's ROI 2025 report — 20-year expected net present value of attending each school, after subtracting the full cost of attendance.
A few things should jump out at you right away.
Pattern 1: STEM-only schools dominate the top — and most of them are tiny
Five of the top six are heavily or exclusively STEM. Two of them — Harvey Mudd (~900 undergrads) and Caltech (~970 undergrads) — could each fit inside a single dorm at a state flagship. Olin College of Engineering, which most high-schoolers have never heard of, has ~400 undergrads total and posts a 20-year NPV of $1.7 million — tied with Caltech and ahead of Princeton.
This isn't a coincidence. The Bureau of Labor Statistics' 2026 outlook still has software developer, data scientist, and operations research analyst as three of the fastest-growing six-figure occupations in the country (BLS Occupational Outlook Handbook). Schools that funnel 80–100% of their graduates into those occupations are going to dominate any median-salary ranking, by construction.
What's interesting is the concentration effect. At a 7,000-person school where 20% major in CS, only 1,400 graduates are in the high-paying track. At Caltech, where 100% of students are in STEM, every single graduate is. That's why Caltech posts a $1.7M NPV at roughly 1/22nd the size of Georgia Tech.
The student-life trade-off is real, though, and nobody tells you this when you're applying:
- Harvey Mudd's humanities, social sciences, and arts (HSA) graduation requirement is famously punishing for a STEM kid who just wants to grind problem sets.
- Caltech's undergrad social scene is sometimes described as "basically grad school with worse food" by current students on r/caltech.
- Olin is so small that if you don't click with the culture, there is nowhere else to go — there are no fraternities, sports teams, or escape valves.
The earnings ceiling is the highest in the country. The fit risk is also the highest in the country.
Pattern 2: Some Ivies are surprisingly mid on ROI
Look at the table again. Of the eight Ivy League schools, only three appear on the top-15 ROI list: Princeton (#5), Penn (#7), and that's it from this specific Georgetown CEW cut. Harvard, Yale, Columbia, Brown, Dartmouth, and Cornell all post strong absolute mid-career salaries — but the return on tuition spent is lower than Caltech, Harvey Mudd, Olin, and Babson because of the higher net price + the wider distribution of majors (English and political science depress the average even at top schools).
The PayScale mid-career rankings reflect this. At 10+ years of experience, MIT ($196,900) and Princeton ($194,100) lead, but U.S. Naval Academy ($187,800), Harvey Mudd ($185,900), and Babson ($181,400) round out the top 5 — ahead of every other Ivy (Harvey Mudd's official announcement summarizing the PayScale data).
The Service Academy point is worth pausing on: the U.S. Naval Academy charges $0 tuition (you pay in service years), and its graduates hit a $187,800 mid-career median. The U.S. Merchant Marine Academy ($0 net price) posts $103,000 at the early-career stage. If you can stomach the service commitment, the ROI math is unbeatable.
Pattern 3: The "schools you've never heard of" tier is real
This is the part of the data students miss the most. There are four schools on this list with <5% national name recognition that out-earn or match most Ivies on lifetime ROI:
- Olin College of Engineering (Needham, MA): 400 undergrads, 21.66% acceptance rate, $1.7M NPV, 89% placed within 6 months of graduation. Founded in 1997 as an experimental engineering school where every student graduates with an engineering degree and a no-grades-in-some-courses pedagogy. Apple, Johns Hopkins, and Procter & Gamble are top employers per their class-of-2023 outcomes.
- Stevens Institute of Technology (Hoboken, NJ): 4,000 undergrads, 47.58% acceptance, $84,800 average starting salary, 96.8% placed within 6 months. Effectively "MIT-lite if you're priced out of MIT and want to work in NYC" — three-quarters of grads stay in the NYC metro.
- Bentley University (Waltham, MA): 4,200 undergrads, 45.05% acceptance, $72,000 median starting, 99% placed. Business-focused, ranked #1 in career services by Princeton Review last year, with the bonus of being commutable to Boston.
- Missouri University of Science and Technology (Rolla, MO): One of the WSJ's top 10 colleges for salary impact, adding $59,624 to graduates' salaries at a $13,773 net price. Pays itself back in 11 months.
These four schools have acceptance rates 5–10x higher than the Ivies but post salary and placement outcomes comparable to (or better than) most of the Ivy 8 on a 20-year horizon. For a student who didn't land Harvard EA, this is the data they should actually be looking at.
Pattern 4: Business-focused outperformers — the Babson wedge
The Harvey Mudd / Caltech / Olin / MIT cluster gets all the press, but there's a parallel cluster nobody talks about: business and entrepreneurship schools that out-ROI most non-Ivy schools by leaning hard into one thing.
Babson College (Wellesley, MA) is the headline example. 3% of its students major in STEM. Its entire pedagogy is built around a single year-long course called Foundations of Management and Entrepreneurship where freshmen launch and run real businesses. The class of 2024 reported its highest-ever average starting salary at $77,681, and the school's PayScale mid-career figure is $181,400 — fifth in the country, ahead of every Ivy except Princeton.
Claremont McKenna College (the same Claremont Colleges consortium as Harvey Mudd) is the other case study. 22% STEM, 9.59% acceptance rate, $80,000 median starting salary for the class of 2024 with JPMorgan, Deloitte, and Microsoft as the top employers. Effectively a feeder school to consulting, banking, and tech strategy roles.
The lesson: you don't have to go pure-STEM to land in the high-ROI tier. You just have to go concentrated — a school that does one thing extremely well and feeds the same recruiting funnel every year.
Pattern 5: Size kills ROI more than people think
The user-facing version of this insight: Georgia Tech is one of the best public universities in the country, with 53,000 students, $13K net price, and a $1.4M NPV — and it still ranks 11th on this list. Caltech, at 1/22nd the size, ranks 3rd.
Why? At a school with 50,000 students, the recruiting and career-services machinery has to operate at industrial scale. Career centers are stretched. The professor-to-student mentoring relationship that lands a Caltech junior an unposted Lockheed internship is much harder to replicate when the cohort is 13,000 freshmen.
This is also why graduate-majority schools (Stanford, Penn, Georgia Tech) often have a different vibe than their undergraduate brochures suggest. At Stanford, graduate students outnumber undergrads ~10,700 to ~7,900. At Georgia Tech, the ratio is even more graduate-heavy: ~33,800 graduate to ~19,200 undergraduate. If you're an undergrad picking between a Stanford and a Harvey Mudd, you're comparing "spend 4 years on a campus where you're a numerical minority" vs. "spend 4 years on a campus that exists entirely for you."
Neither answer is wrong. But the comparison matters for fit and outcomes.
The acceptance-rate gradient is also a recruiting gradient
Here's where this gets practical for high-schoolers. Look at acceptance rate vs. starting salary:
- Caltech: 3.14% accept rate → ~$110K+ median starting
- Stanford: 3.61% → ~$100K+
- MIT: 4.52% → ~$110K
- Princeton: 4.62% → ~$95K+
- Harvey Mudd: 12.66% → $117,500 ✨
- Carnegie Mellon: 11.66% → ~$110K+ (CS / School of Computer Science specifically)
- Babson: 17.00% → $77K
- Olin: 21.66% → ~$85K+
- Bentley: 45.05% → $72K
- Stevens: 47.58% → ~$85K
Two things jump out:
- The top of the salary curve is much flatter than the top of the selectivity curve. Going from a 3% acceptance rate to a 21% acceptance rate (Caltech → Olin) only costs you ~$25K in median starting salary. Going from 21% to 47% (Olin → Bentley) only costs you another $10–15K.
- Selectivity alone does not predict salary. Harvey Mudd has a 4x higher acceptance rate than Caltech and a 3x higher rate than Stanford — but it has the highest published class-of-2024 median starting salary in the U.S.
Why does this matter? Because most students apply to colleges on a prestige axis and then compare salaries afterwards. The right move is to look at salary outcomes first, then back into the acceptance rate. There are at least six schools on this list with single-digit-to-low-double-digit acceptance rates where the salary outcomes match or beat the Ivies. The applicant pool that beats themselves up over a Harvard rejection often ignores Harvey Mudd entirely.
What this means if you're applying
Three honest takeaways from the data:
1. The "fit + concentration" pair beats brand. Caltech, Harvey Mudd, Olin, Babson, and Claremont McKenna are all concentrated, fit-sensitive schools. They reward students who know what they want and punish students who don't. If you're 90% sure you want to be a software engineer, Harvey Mudd or Olin will out-recruit you into an internship faster than an Ivy with a more diffuse major mix. If you're 50% sure, the bigger school with optionality matters more.
2. Net price >> sticker price. MIT's sticker is $82,676, but the average net price after aid is $19,813. Princeton's net is $10,555 because of full-ride income thresholds. Babson's is $38,876. The lower-net schools (Princeton, Stanford, Georgia Tech) often beat the higher-net schools on ROI even when their salary outcomes are slightly lower. Always run the net-price calculator before judging by sticker.
3. The Service Academies are the most underrated ROI play. Naval, Air Force, West Point, Coast Guard, Merchant Marine — all $0 net cost with mid-career salaries north of $130K (and Naval Academy at $187,800). If you can do the service commitment, the ROI math is uncatchable.
The bridge: your school is a 30-second filter, your resume is the next 4 years
Here's the part most career advice ignores.
Every recruiter we talk to — at Microsoft, Google, JPMorgan, McKinsey, Pfizer — uses the school name on your resume as a 30-second filter, not a decision. Going to MIT gets your resume opened. It doesn't get you the interview. Going to Bentley also gets your resume opened (your career center has spent 20 years building relationships with these recruiters), but the bullet quality matters more because the brand pull is less.
What this means in practical terms:
- If you're at one of the top 5 on this list (MIT, Harvey Mudd, Caltech, Olin, Princeton), your resume will open. Your interview conversion depends on whether the bullets actually demonstrate impact — see our breakdown of achievement bullets that prove impact.
- If you're at one of the schools 6–13 (Stanford, Penn, CMU, Babson, Bentley, Georgia Tech, Claremont McKenna, Stevens), your career center has done the hard work of relationship-building with recruiters. Your bullets need to be as tight as the top-5 resumes, but you also need to tailor each one to the JD because the brand pull is thinner.
- If you're at a school that isn't on this list (most students, including the writer of this post), your resume is doing 100% of the work. The recruiter doesn't have a 20-year relationship with your career center. The school name buys you 5 seconds, not 30. Every bullet has to earn its place — see 12 resume mistakes recruiters reject and ATS resume keywords 2026 for the specific patterns to fix first.
In all three cases, the resume is the deliverable. The school name is just how long you get to make your case.
The most common ResuAI use case we see from college students is exactly this: senior year, two months before recruiting season, comparing their resume against the JDs they want to apply to and realizing the bullets don't match the language of the JD. That's a fixable problem — usually in one focused afternoon. Run yours through the analyzer with the JD pasted in: you'll see the exact keywords missing, the bullets that read as job-description-paraphrasing rather than impact, and the specific rewrites that fix each one.
For LinkedIn, the equivalent is the LinkedIn optimizer — paste your target role JD, upload your LinkedIn PDF, get a per-section score and a rewritten headline + About in your voice. The recruiters at MIT-tier companies search LinkedIn before they search the applicant pile, and a LinkedIn that doesn't surface in the right searches loses you opportunities that never come back.
TL;DR
The 2026 data is unambiguous: the colleges that produce the highest-earning graduates in the United States are disproportionately small, concentrated, and STEM-heavy — with a parallel cluster of business-focused outperformers (Babson, Bentley, Claremont McKenna).
- Highest published class-of-2024 median starting salary in the U.S.: Harvey Mudd at $117,500.
- Highest mid-career median salary: MIT at $196,900, Princeton at $194,100, U.S. Naval Academy at $187,800.
- Highest 20-year NPV: MIT at $1,881,000, Harvey Mudd at $1,819,000, Caltech at $1,719,000.
- Best ROI per dollar of net price: Princeton (5-month payback), Georgia Tech (8-month), MIT (8-month), Stanford (6-month).
- Best "school you've never heard of": Olin College of Engineering ($1.7M NPV, 21.66% acceptance), Stevens Institute of Technology (96.8% placement, 47.58% acceptance).
If you're picking schools: optimize for concentration and fit, not pure brand. If you're already in school: your resume is doing more work than the brand on it — get the bullets right before recruiting season opens.
Sources for the salary data in this post:
- Town & Country: 15 Colleges With the Best ROI 2025 — class-of-2024 senior outcomes citations
- Georgetown CEW: Ranking 4,600 Colleges by ROI (2025) — 20-year NPV figures
- PayScale College Salary Report — early-career and mid-career medians
- Harvey Mudd: PayScale Ranks Harvey Mudd No. 1 for Early Career Salaries
- CNBC / Wall Street Journal: The 10 best U.S. colleges for graduate salary impact
- College Kickstart: Class of 2030 Admission Results — most recent acceptance rates

Written by
ResuAI Editorial
ResuAI's in-house editorial team reads 200+ job descriptions a week to keep our analyzer (and these guides) sharp.
We're the small team that builds, breaks, and re-tunes the ATS scoring engine, the resume builder templates, and the analyzer's bullet rewrites. Everything we publish is grounded in what real recruiters and ATS systems actually do today -- not the conventional wisdom that's been recycled since 2014.
Try this on your own resume
Run your resume + the JD through the analyzer for a match score, missing keywords, and bullet rewrites.
Keep reading
More from the career hub
9 min read
The Realistic Job Search Timeline (How Long It Actually Takes in 2026)
United Way's 2026 survey of 1,000 job seekers: 6.6 months average, 62.6 applications, 94% experience ghosting. Here's the timeline by experience level — and how to pace yourself.
9 min read
Cold Message Templates That Actually Get Replies (Recruiter, Hiring Manager, Referral)
10 cold-message templates that move LinkedIn reply rates from <5% to 25-40%. Per benchmark data: cold email = 3%, InMail = 18-25%, under-400-char messages = 22% lift.
10 min read
The LinkedIn Profile Checklist That Actually Gets Recruiters to DM You (2026)
95% of recruiters use LinkedIn. They spend 6 seconds on your profile. Here are the 14 specific changes — in priority order — that move you from invisible to inbound.