Data science sits at the intersection of two things Harvard does exceptionally well: statistics (through the Department of Statistics) and applied computation (through SEAS), and the SM in Data Science is Harvard's answer to the surge in demand for graduates who can move fluidly between machine learning, statistical inference, and real-world data infrastructure. It's explicitly listed among SEAS's popular graduate programs, reflecting how quickly it has grown from a newer offering into one of the school's most in-demand master's degrees. For applicants comparing it to Stanford's MS in Statistics/Data Science or MIT's data-focused tracks, Harvard's version offers a genuinely interdisciplinary structure — housed jointly across SEAS and Statistics — plus Boston's dense biotech, health-data, and finance industry base as a built-in recruiting pipeline.
This guide covers the acceptance rate reality, cost, admission requirements, how to get in, English-test rules, and outcomes specifically for Harvard's SM in Data Science.
Acceptance Rate: What Harvard Actually Publishes
Harvard does not publish a program-specific acceptance rate for the SM in Data Science. Here's the honest picture, built from what Harvard does disclose:
- SEAS (via GSAS) admits an estimated 5-8% of applicants across its graduate programs. Data Science is one of SEAS's most popular and fastest-growing master's programs, listed explicitly alongside SM in Computer Science and SM in Computational Science and Engineering in Harvard's own graduate program list — so it competes for a similarly deep, strong applicant pool, and its actual admit rate could plausibly sit at or below the SEAS-wide band.
- GSAS overall reports roughly 3.1%, but this blends all GSAS departments from data science to the humanities, and is not a reliable substitute for a data-science-specific number, since applicant pool size and quality vary enormously by department.
- There is no separate undergraduate "data science acceptance rate" at Harvard College, since admission is to the College as a whole (~3.6-4.2% overall), not to a specific concentration; Harvard College does not have a standalone data science concentration in the way SEAS's graduate program does — undergraduates typically pursue this interest through Statistics, Computer Science, or Applied Math concentrations instead.
Bottom line: treat any number below roughly 8% as an informed estimate rather than a Harvard-published fact, and prepare an application as competitive as you would for any single-digit-admit-rate graduate program in the country.
Fees and Cost of Attendance
The SM in Data Science follows Harvard's standard SEAS/GSAS graduate tuition schedule:
| Cost Item | Amount (USD/year) |
|---|---|
| Tuition | $57,328 |
| Living expenses (Cambridge/Boston) | $35,150 |
| Total estimated cost of attendance | $98,621 |
Most SM in Data Science programs run one to two years depending on the track and whether a student enters with a strong quantitative background; budgeting $100,000-$200,000 total (tuition plus living costs, before any assistantship or funding) is realistic. The graduate application fee is $105, with fee waivers available for demonstrated financial hardship.
As with all Harvard programs, tuition is identical for domestic and international students — there is no international surcharge.
Admission Requirements
The SM in Data Science is administered through GSAS/SEAS with the standard application components:
- Academic transcripts from every post-secondary institution attended
- Statement of Purpose, ideally identifying a specific focus (statistical methodology, machine learning systems, applied domains like computational biology or economics)
- Three letters of recommendation, preferably from people who can evaluate quantitative and/or research ability directly
- Resume/CV
- GRE: not required for most GSAS programs in recent cycles, but this is department-specific and worth reconfirming before applying
- TOEFL or IELTS scores, required unless a waiver applies (see below)
- $105 application fee or an approved fee waiver
Harvard does not publish a minimum GPA, but competitive Data Science SM applicants typically hold strong quantitative degrees (statistics, computer science, applied math, economics, engineering) with substantial coursework in linear algebra, probability/statistics, and programming, plus evidence of applied project or research work involving real datasets. The regular GSAS deadline is typically December 1, with decisions generally out by February-March.
How to Get In
-
Show quantitative depth, not just coding ability. Data Science SM applicants are evaluated on statistical reasoning as much as programming — a strong applicant demonstrates comfort with probability theory and inference, not just familiarity with pandas or scikit-learn.
-
Anchor your statement in a specific applied problem. The strongest applications describe a concrete data problem the applicant has worked on (a research project, a Kaggle competition taken seriously, a work-based analytics initiative) rather than a general interest in "data science as a field."
-
Leverage the interdisciplinary structure. Because the program straddles SEAS and Statistics, applicants coming from adjacent fields (economics, biology, public health, social science) with strong quantitative training are genuinely competitive — don't assume only CS undergrads are a fit, but do make the quantitative rigor of your background explicit.
-
Recommendation letters should speak to analytical rigor. A reference who supervised you doing real statistical or ML work (not just general professionalism) carries disproportionate weight for this specific program.
-
If GRE is optional, submit only if it's a clear positive. A near-perfect quantitative GRE score can help differentiate an application from a less familiar academic system, but a mediocre score adds nothing and is better omitted where the test is optional.
English Proficiency Requirement (IELTS/TOEFL)
The SM in Data Science follows GSAS's university-wide graduate English-proficiency policy (Harvard does not set a separate threshold per SEAS program):
- IELTS: minimum 7.0 overall, recommended 7.5; GSAS requires a minimum Speaking sub-score of 6.5
- TOEFL: minimum 95 overall, recommended 104; GSAS requires a minimum Speaking sub-score around 23/30 (4.5 on the redesigned scale for tests from January 21, 2026 onward)
- Waiver: granted only for applicants whose undergraduate (not graduate) degree was earned at an English-medium institution; GSAS does not accept Duolingo or PTE
Given how quantitatively selective this program is, a borderline English score can become an unnecessary point of friction in an otherwise strong file — targeting the recommended 7.5 IELTS / 104 TOEFL, with particular attention to the Speaking sub-score minimum, is the safer approach. For full details on Harvard's English-test policy across schools, see Harvard University IELTS requirements and Harvard University TOEFL requirements.
Career Outcomes
Harvard does not publish outcomes data broken out specifically for the SM in Data Science, but Harvard's graduate-wide figures are a reasonable proxy, and data science compensation specifically tends to sit at or above these blended averages given strong demand:
- Graduate employment rate after 6 months: ~90% (university-wide graduate figure)
- Average graduate starting salary: ~$184,500 (university-wide graduate figure; data science and machine-learning roles at top employers frequently exceed this, particularly in tech and quant finance)
- Top recruiters (university-wide graduate): McKinsey & Company, Bain & Company, Boston Consulting Group, Goldman Sachs, Blackstone, and private equity/venture capital firms — Data Science SM graduates specifically are also heavily recruited into applied ML and analytics roles at major tech companies, biotech/health-data firms (leveraging Boston's life-sciences cluster), and quantitative finance shops
The combination of Harvard's brand, a genuinely interdisciplinary curriculum, and Boston's dense data-driven industry base (health tech, biotech, finance, and a growing AI research scene anchored partly by Harvard's own Kempner Institute) gives SM in Data Science graduates unusually broad exit options relative to a narrower, purely technical master's degree.
FAQ
Does Harvard publish a separate acceptance rate for the SM in Data Science? No. There is no published program-specific figure. The best available proxies are the SEAS/GSAS graduate admit rate (roughly 5-8%) and GSAS's overall 3.1% rate, though Data Science's popularity as one of SEAS's fastest-growing programs suggests real selectivity at or beyond the SEAS-wide band.
Is the GRE required for Harvard's SM in Data Science? Most GSAS programs, including SEAS's data science track, do not require the GRE in recent cycles, but confirm current policy directly with the department before applying, since this can change.
What background do I need to be competitive for this program? A strong quantitative undergraduate degree — statistics, computer science, applied math, economics, or engineering — with solid coursework in probability, statistics, linear algebra, and programming, plus demonstrated applied project or research experience with real data.
What IELTS or TOEFL score do I need? GSAS's minimum is IELTS 7.0 (Speaking at least 6.5) or TOEFL 95 (Speaking around 23/30), but aiming for the recommended IELTS 7.5 or TOEFL 104 is the safer target for this competitive program.
How much does the SM in Data Science cost at Harvard? Tuition is $57,328/year, with total cost of attendance (including living expenses) around $98,621/year — expect $100,000-$200,000 total depending on program length, before any funding.
Can non-CS undergraduates apply to Harvard's Data Science SM? Yes. Because the program is interdisciplinary across SEAS and Statistics, applicants from economics, biology, public health, or social science backgrounds with strong quantitative training are genuinely competitive, provided their statistical and programming foundations are solid.