Skip to main content

    GradsMatch Research·Issue 01·Cost & affordability

    Master's tuition varies 100-fold across the developed world: a 35,000-program analysis

    Median published tuition for a US master's program is $61,770. The most common published tuition for a German master's is $0 - public universities outside Baden-Württemberg charge no tuition, only a Semesterbeitrag of €100-€400. The gap holds across most fields.

    All figures are sticker tuition, before aid. Net price is a separate question we don't try to estimate here.

    US median tuition

    $62K

    n = 9,508

    European median

    <$10K

    DE, FR, IT, ES, BE, AT

    Programs analyzed

    33,633

    After outlier trim

    Countries covered

    27

    ≥25 programs each

    Key findings

    1. 01

      The outlier

      The US is in a class of its own.

      Median total published tuition for a US master's program is $61,770 (n = 9,508). That's substantially higher than every other country in the dataset, including all other G7 economies. Australia is the only other country whose median crosses $60K.

    2. 02

      The contrast

      Continental Europe stays cheap, even at top schools.

      Germany, France, Italy, Belgium, Austria, and Spain all have median master's tuition below $15K. Several publish credible programs under $2K once you filter to public universities. (See methodology - some of those low values reflect EU-citizen rates that don't apply to international applicants.)

    3. 03

      The field

      The gap is structural, not field-specific.

      In Computer Science specifically: US median $62K, Germany $0, Italy $7.0K. The pattern repeats in business, engineering, and the social sciences. The full heatmap is below.

    4. 04

      The fees

      Application fees mirror the same divide.

      85% of US programs in our dataset charge an application fee (median $85, max $1,000 at six institutions including Purdue, Boston University, Notre Dame, and NC A&T State). 77% of German programs in our dataset charge nothing to apply, as do 94% of French programs. The cumulative cost of applying scales with both the number and the country.

    Fig. 1

    Median master's tuition, all fields

    Each bar is the median total published tuition for a master's program in that country, in USD, across all fields. Tooltips show the 25-75 percentile range and sample size. Countries with fewer than 25 programs in our database are excluded so a single high-cost or low-cost outlier doesn't drive the result.

    Source: GradsMatch database, 2026-05-07. Median total published tuition; 99% inclusion window applied (see methodology).

    Why is the gap so large?

    US master's tuition is high for the same reason undergraduate tuition is high: the country combines unusually expensive private universities with state schools that charge non-residents (and especially international students) closer to private-school rates. The price floor is set by the most expensive participants, not the cheapest.

    In most of continental Europe, by contrast, public universities still charge nominal registration fees regardless of nationality, or at most a few thousand euros per year. That keeps the median low even at top-ranked institutions. Germany's technische Universitäten, France's grandes écoles, Italy's research universities, and the Netherlands' research-oriented programs all sit in this bracket. The premium you pay at private programs in those countries (HEC Paris, IESE, Bocconi) is real but rarer.

    The English-speaking outliers - UK, Australia, Canada, New Zealand, Ireland - are increasingly closer to the US than to Europe. Their pricing models emerged from the 2010s push to recruit fee-paying international students, and they reflect that. UK postgraduate tuition for international students has risen roughly 40% in real terms over the past decade according to Universities UK; Australia's pattern is similar.

    Fig. 2

    The shape of the distribution, not just its height

    Each dot is one Computer Science master's program in our database; rows are countries, ordered by median (most expensive on top). The bar chart above showed that medians differ. This shows the dispersion: where each country's programs cluster, where they have a tail, and where the cheap US online options actually overlap with mid-range European programs.

    Source: GradsMatch database, 2026-05-07. One dot per CS program, with a small jitter so overlapping points are visible. Red dashed line: US median. Green dashed line: highest median among Germany / France / Italy / Spain / Belgium / Austria. Up to 200 programs sampled per country.

    Wide tail

    The US distribution stretches from roughly $20K (online MS programs at state schools) to nearly $200K. The gap between cheapest and most expensive US CS master's is larger than the entire span of any European country's CS market.

    Tight clusters

    Germany, Italy, France, Austria, and Spain all show tight clusters near zero. Public universities dominate the sample; the private outliers that exist (e.g. Bocconi, IE, HEC adjuncts) are visible as the lone dots well above the cluster.

    Where they overlap

    The cheap US online options (~$20-30K) sit in the same range as many UK, Dutch, and Belgian master's. "US is more expensive" is true on average but breaks down once you compare specific programs - which is the actual decision applicants face.

    Fig. 3

    Median tuition by field and country

    The pattern repeats across fields. Cells with fewer than 5 programs are shown as a dash, not because the field doesn't exist there, but because we don't yet have enough programs to compute a credible median. Hover any cell to see n.

    FieldAustraliaUnited StatesUnited KingdomNetherlandsSwedenCanadaIrelandChinaFranceItalyGermanyJapan
    Computer Science$62Kn=49$62Kn=351$42Kn=155$53Kn=19$33Kn=27$15Kn=19$33Kn=18$12Kn=59$5.8Kn=10$7.0Kn=16$0n=66$674n=13
    Business / MBA$59Kn=291$66Kn=708$38Kn=669$29Kn=98$30Kn=44$48Kn=122$27Kn=64$13Kn=175$9.3Kn=45$7.6Kn=80$7.1Kn=133$799n=31
    Mechanical Engineering$91Kn=5$58Kn=136$43Kn=59·$31Kn=8$24Kn=17$24Kn=5$11Kn=21·$6.9Kn=17$1.1Kn=9$674n=10
    Electrical Engineering·$62Kn=129$47Kn=67$51Kn=7$35Kn=12$22Kn=25·$12Kn=61$10Kn=5$6.9Kn=24$0n=20$674n=13
    Education$57Kn=22$73Kn=130$37Kn=58$25Kn=9·$31Kn=15$22Kn=5$13Kn=43··$7.1Kn=9$674n=7

    Source: GradsMatch database, 2026-05-07. Cell color intensity reflects median. Hover for n.

    Application fees move with the same gradient

    Tuition is the headline cost, but applying isn't free either. The country distribution of application fees mirrors the tuition distribution: US programs charge a fee 85% of the time at a median $85; six programs in our dataset charge $1,000 (Purdue, Boston University, Notre Dame, North Carolina A&T State). Continental Europe applies for free far more often: 77% of German programs in our dataset and 94% of French programs charge no application fee.

    Caveat: real applicants don't apply uniformly to ten programs at the median fee. Selectivity tier, fee waivers for low-income applicants, and portfolio strategy all materially shift any cumulative-cost figure. Treat the per-program rate as the comparable number across countries.

    CountryPrograms charging a feePrograms with no feeMedian fee% free
    United States9,8481,439$8013%
    United Kingdom1,9894,543$11570%
    Canada1,615310$9516%
    Australia1,0841,002$10748%
    Germany6721,936$8874%
    France72670$11090%
    Netherlands784329$11830%

    Application-fee data is sourced from the same official program pages as tuition data. A program counted as "free" either explicitly waives the fee or doesn't charge one in the listed admission cycle.

    What the headline number doesn't capture

    Tuition is one part of a master's degree's true cost. Three things this analysis deliberately doesn't try to estimate:

    • Cost of living. Living for two years in Zurich, London, or New York isn't comparable to two years in Berlin, Lyon, or Tokyo. Where you'd save on tuition you may spend on rent.
    • Funding and assistantships. Many US PhD-track programs and a meaningful share of master's programs offer tuition waivers and stipends. Our dataset records sticker tuition; we have a separate funded master's guide that surfaces the exceptions.
    • Post-graduation outcomes. Median post-graduation salary varies enormously across countries, both because of labor-market wages and because of what visas allow you to take after graduation. The US STEM-OPT extension is a real economic factor for many international applicants weighing US programs against cheaper alternatives.

    None of those caveats erase the headline finding - sticker tuition really is an order of magnitude apart between the US and most peer countries - but they're worth holding in mind before drawing personal conclusions about where to apply.

    Methodology

    Data source

    All figures are computed from the GradsMatch database, which currently tracks 40,374 master's programs across 3,272 universities. Programs are extracted from official university websites. Each program record links back to a primary source URL and carries a "last verified" date. Full data collection details are at /methodology.

    Currency normalization

    Tuition is recorded in the program's native currency (USD, GBP, EUR, JPY, etc.) and converted to USD using daily FX rates. We preserve both the native value and the USD value; the analysis uses USD throughout, but cards on individual program pages display the native amount so applicants don't compare a converted number against their bank balance.

    Outlier handling

    Two filters keep the analysis robust against extraction errors:

    • Inclusion window. Programs with USD-equivalent total tuition above $200,000 are excluded; this catches multi-year cumulative figures and OCR errors. We do not apply a low-end floor: $0 is a real, common figure (most German public universities, most Brazilian federal universities, several Nordic programs for EU citizens), and excluding it would silently misstate the cross-country comparison.
    • Minimum sample size. A country needs at least 25 programs in our database to appear in the all-fields chart, and at least 5 programs in a given field to appear in the field-specific charts. Below that threshold we don't think the median is informative.

    Germany: tuition defaults

    Germany is the largest single source of programs in our database where the official program page reads "no tuition fees" rather than publishing a number. To avoid silently dropping ~1,500 German master's programs from the analysis, we apply a small set of policy-grounded defaults sourced from DAAD and verified against each state's higher-education ministry:

    • $0 for non-EU master's at public universities in 14 of 16 German states (Berlin, Hamburg, NRW, Saxony, Lower Saxony, Hesse, etc.). Tuition is statutorily zero; students pay only a Semesterbeitrag of €100-€400 per term, which we do not include.
    • $3,300/year (€1,500/semester × 2) for non-EU master's at Baden-Württemberg public universities (Heidelberg, KIT, Freiburg, Stuttgart, Mannheim, Tübingen, Hohenheim, Ulm, Konstanz). This rate is set by state law.
    • $11,000/year (€5,000/semester midpoint × 2) for non-EU master's at TUM, the only Bavarian public to use its 2023 fee authority. LMU and other Bavarian publics remain $0.
    • Per-program extraction for German private universities (Frankfurt School, ESMT, Hertie School, EBS, WHU, Bucerius, Jacobs, Witten/Herdecke), MBA and executive programs at any university, and any program whose URL we could verify a specific published fee against.

    Each backfilled row carries a provenance tag (tuition_assumption_source) so the assumption is auditable per program. The headline finding - that the German median is dramatically below the US - is unaffected by reasonable variation in these defaults because the public-uni share dominates the German master's market.

    Statistic of choice

    We lead with the median rather than the mean because tuition distributions are right-skewed (a handful of expensive private programs pull the mean upward). Means and 25-75 percentile ranges are available in the chart tooltips for readers who want them.

    Field classification

    Programs are classified into ~70 internal "buckets" (e.g. CS_IT, BUSINESS_MGMT, ENG_MECH_AERO) using a custom matcher that reads program names and descriptions. The taxonomy mostly mirrors NCES CIP codes but is broader at the leaves. The field labels in this article are human-readable mappings of those buckets.

    EU vs international student rates

    Some countries - France, Belgium, Germany, the Nordics - charge dramatically different tuition to EU/EEA citizens vs international students from outside the bloc. France's public master's tuition is roughly €243/year for EU citizens but €3,770/year for international students; Belgium's range is similar; Sweden and Finland charge nothing for EU students but €10,000-20,000/year for non-EU.

    GradsMatch records whatever tuition figure the program publishes most prominently on its public-facing page. That's usually the standard rate but can be the EU rate at universities that bury the international figure deeper in their fee schedule. Where a program publishes only the EU rate, that's what we have. The absolute lowest tuition values in our French and Belgian data (under $1,000) correspond to public-university EU-citizen pricing - they're real numbers, but international applicants from outside the EU should expect to pay 5-15× those figures at the same schools.

    This is the largest single source of imprecision in the cross-country comparison. The headline finding (that the US is far more expensive than continental Europe) holds even after correcting for EU-vs-international rates - international rates in France/Germany/Belgium typically max out around $15,000/year, still well below the US median - but specific country-level medians shift if you condition on student citizenship.

    Refresh cadence

    Tuition data is re-verified continuously against official university pages. Each chart on this page is regenerated every time GradsMatch deploys, so the numbers in the charts and the numbers in the prose stay synchronized. The "last updated" date in the byline reflects the most recent build.

    Download & cite

    The country-level and field-by-country medians, percentiles, and sample sizes used in this analysis are available as a CSV. Released under CC BY 4.0 - reuse freely with attribution.

    Download CSVFrozen snapshot · v2026.05.06 · 2026-05-07

    Suggested citation

    Antoine P. (2026). Master's tuition varies 100-fold across the
    developed world: a 35,000-program analysis. GradsMatch Research,
    Issue 01 (v2026.05.06). https://gradsmatch.com/research/master-tuition-by-country

    For derived cuts that aren't in the public CSV (program-level records, custom field definitions, school-tier breakouts, year-over-year if available), email antoine@gradsmatch.com - usually a same-day turnaround.

    Explore the underlying programs

    Related