The Full R2 Universities List and RCU Research Colleges

An R2 university clears two bars at once. Carnegie's 2025 rule is that the institution spends at least $5 million a year on research and development, as reported to the National Science Foundation's HERD survey, and awards at least 20 research doctorates a year, as reported to IPEDS. 139 institutions met both in 2025. The R1 band above uses the same two measures with the numbers raised to $50 million and 70 doctorates, and the state-by-state R1 universities list covers those 187 schools.

Those thresholds are lower than they sound in practice. The median R2 in this file spends $26.4M a year and awards 42 research doctorates. That is a real graduate school with externally funded projects, and it is also about half the $50 million the R1 money threshold asks for.

What an R2 reliably has that an RCU usually does not: doctoral programmes established in more than one field, since holding the designation already requires at least 20 research doctorates a year, a graduate school with its own dean and funding lines, and core facilities such as shared instrumentation, a sequencing or microscopy centre, or a high performance computing cluster, because a $20 million research operation cannot run without them. What R2 does not tell you is anything about class size, advising, or whether an undergraduate can get into a lab.

RCU is new in 2025 and it exists to fix a specific blind spot. The old classification could only see research through doctoral production, so a college that ran a serious externally funded research operation but granted few or no PhDs was invisible. Carnegie's rule for RCU is one number: at least $2.5 million a year in research spending. Institutions already in R1 or R2 are excluded. 216 institutions carry the designation.

The consequence shows up immediately in the data. 126 of the 216 RCU institutions award no research doctorates at all, and 51 carry a Baccalaureate Colleges classification, meaning a bachelor's degree, not a doctorate, is their predominant award. The median RCU spends $5.5M a year. This is a band of liberal arts colleges, regional master's universities, service academies, tribal colleges and standalone medical and research centres.

Two RCU schools and what they actually research

South Dakota School of Mines and Technology (RCU, public, South Dakota) spends $20.3M a year, the largest research operation of any RCU that is within reach of R2 on both axes. Its named centres include the Governor's Research Center for Electrochemical Energy Storage, the NSF industry-university centre for solid-state electric power storage, and a composites and biomaterials manufacturing centre. It awards 18.3 research doctorates a year, which is what keeps it out of R2.

Diné College (RCU, tribal college, Arizona) spends $3.1M a year and awards no research doctorates. Its research office runs the Diné Policy Institute, a Navajo language immersion institute, its own institutional review board, and NSF and NIH-funded summer research programmes for undergraduates. Under the pre-2025 rules a college like this registered as having no research at all.

The extreme case in the file is The University of Texas MD Anderson Cancer Center: $1.26B a year in research spending, the largest figure of any institution in either band, and an RCU because it awards no research doctorates in its own name. If you took the money alone as the signal, you would rank it above most of the R1 list.

Run the two thresholds separately across the file and the asymmetry is stark.

Among the 139 R2 institutions, 31 already spend more than $50 million a year, the R1 money threshold. Every one of those 31 is held out of R1 by the doctorate rule alone. 30 more already award 70 or more research doctorates and are held out by money alone. The remaining 78 fall short on both.

Among the 216 RCU institutions the gap is wider still: 116 of 216 already spend more than the $5 million R2 threshold, and every one of them is stopped by the doctorate count. Only 8 RCU institutions award 20 or more research doctorates.

The reason is structural. Research spending can jump in one budget year when a large federal grant lands. A doctoral programme cannot: it needs a state or system authorisation, an accreditor sign-off, faculty with the standing to chair committees, and then five to seven years before the first cohort defends. Money is a tap. Doctorates are a pipeline with a fixed transit time.

The honest answer is that the designation is a proxy for how crowded the lab bench is, and that is the only thing it is a good proxy for.

Work the arithmetic on a median R2: 42 research doctorates a year across the entire institution. Spread across thirty or so departments, a given department graduates one or two PhDs a year, which means a lab typically holds two or three doctoral students at a time. When a professor there needs a pair of hands for a data-collection run, an undergraduate is a realistic answer rather than a courtesy. At an RCU with no doctoral programme at all the effect is stronger: the money is real, and there is nobody else in the queue.

The trade-offs are just as real, and you should hear them stated plainly:

1. Narrower subject coverage. An R2 spending $26.4M is concentrated in a few fields. If the funding is in nursing and civil engineering and you want particle physics, the institutional number tells you nothing useful.
2. Thinner equipment. Core facilities exist, but the instrument you want may be one machine with a waiting list, not four.
3. Weaker onward signalling for a research PhD. A letter from a well-known specialist carries further than a letter from a good teacher, and specialists cluster at R1. This matters for PhD admissions and much less for anything else.
4. The label is not a quality mark. Neither threshold measures teaching, advising, graduation rates or debt. Compare those separately, starting with what net price actually means on an offer letter.

Before you weight any of this, check that the subject is right for you at all. Field choice moves outcomes further than tier does, which is the point of the workload comparison across college majors.

Here the designation carries more weight, because it maps onto money you can actually be paid.

A thesis track needs a supervisor with an active project and a committee that can examine it. R2 institutions have both by construction: they run doctoral programmes, so the committee structure already exists and a master's thesis slots into it. At an RCU with no doctoral programme, thesis tracks exist in some departments and not others, and you have to check the department page rather than the institution.

Assistantship depth follows the research dollars, not the tier label. An R2 spending $60 million has grant lines that can carry a master's RA at a stipend plus tuition remission. An RCU spending $3 million usually cannot, and will offer hourly work instead. This is why the spending column on this page is worth more to you than the letter.

The practical conclusion most applicants under-use: a funded RA place at an R2 beats an unfunded seat at an R1. Two years of a full-price unfunded master's at a well-known R1 can cost more than the salary premium it buys, while a funded R2 place pays you to do the research that a PhD application is actually judged on. Set the offers side by side with a college list builder that keeps cost and fit in the same view before you decide.

One caution: read the funding offer, not the designation. Ask the department how many first-year master's students were funded last year, for how many semesters, and whether the funding is tied to one grant that ends.

Sort either tab by “closest to” and the top of the list is the set of schools where one axis is nearly full and the other is not far behind. Read the ranking as a description of distance, not a prediction. Two cautions apply.

The cycle. Carnegie re-runs the designations every three years and the next release is scheduled for 2028. A school that is 3% short today has one reporting cycle to close the gap, and a school comfortably inside a band can also fall out.

The rule is changing. The 2025 designations used, in Carnegie's words, “the higher of either a three-year average (2021, 2022, 2023) or the most recent single-year data (2023)”, and Carnegie states that “in future releases, the designations will use only a three-year average”. Reproducing that rule against the public data file returns all 542 designations exactly, which lets you see who benefited from it: 22 of the 139 R2 institutions and 25 of the 216 RCU institutions would not clear their own threshold on the three-year average alone. Those rows are tagged “2023 year only” in the explorer. If their 2023 figure was a one-off, they are the schools most likely to lose the label in 2028, not gain a better one.

If a school's designation is part of why you are applying, that tag is the single most decision-relevant thing on this page. For the tier above and the way its threshold was rewritten in 2025, see what R1 classification measures and what it does not.

1. Open the Carnegie institution lookup and search the school by name.
2. Read the row labelled 2025 Research Activity Designation. It will say Research 1, Research 2, Research Colleges and Universities, or nothing at all.
3. For the underlying numbers, download the 2025 Research Activity Designation Public Data File from the designations methodology page. Columns I and M hold the three-year averages; columns H and L hold the 2023 single-year figures.
4. Apply the rule yourself: take the higher of the average and the 2023 figure on each axis, then compare to the thresholds.