Summer programmes split into four types that behave completely differently: free and highly selective research programmes, paid pre-college courses, structured mentorship programmes, and local university placements. The free selective ones are mostly closed to international students and to rising seniors. The paid pre-college ones admit nearly everyone who can pay, and cost between £3,000 and £13,000. Work out which type you are actually looking at before you spend a single evening on an application.
Search "summer programs for high school students" and you will get lists that put a free seven-week research placement at MIT in the same table as a two-week paid course that admits almost everyone who applies. Those two things have nothing in common except the month they happen in.
This guide sorts the field into the four types that actually behave differently, with the real numbers for each.
The four types
| Type | Typical cost | Selectivity | What you leave with |
|---|---|---|---|
| Selective research programme | Free, sometimes with a stipend | High; small cohorts | Original work, a named supervisor, a cohort |
| Pre-college course | £3,000–£13,000 | Low; mostly an academic floor | Course credit, a transcript line |
| Mentorship programme | £1,500–£5,000 | Moderate; applications assessed | A finished paper, often publishable |
| Local university placement | Free | Varies; often informal | Lab experience, a reference |
Almost every disappointment in this area comes from a student applying to one type while expecting the outcomes of another.
Type one: selective research programmes
These are the ones worth structuring a year around. They are free, small, and they put you in a working research group.
The Research Science Institute, run by the Center for Excellence in Education, takes 100 students to MIT each summer at no cost, and for 2027 is expanding to Stanford for a further 50 places. The structure is one week of intensive STEM classes, a five-week research internship, and a final week of presentations. Two things catch applicants out: high school juniors are eligible and seniors are not, and applications for the 2027 programme close at 11:59pm on 11 December 2026 — about seven months before it runs.
SIMR, Stanford's Institutes of Medicine Summer Research Program, runs roughly eight weeks from 8 June to 30 July 2026. Its eligibility rules are stricter than most lists admit: you must be 16 by the start date with no exceptions, you must be a US citizen or permanent resident living in the US, and students within about an hour's drive of the Bay Area get preference. Stipends exist but are limited and needs-based, generally for families with adjusted gross income under $80,000 — not a flat payment to every participant, as aggregator guides often imply.
The Anson L. Clark Scholars Program at Texas Tech is the smallest of the well-known ones: twelve students, seven weeks from 21 June to 6 August 2026, a $750 stipend paid on completion of a successful research report, and no participation fee — room, board and meals are covered. You must be 17 by the start date, graduating in 2026 or 2027, and a US citizen or permanent resident. The page says "NO EXCEPTIONS" in capitals.
The Summer Science Program is the main selective option genuinely open to international students. It admitted around 700 participants in 2026, uses need-blind admission, and offers $3,000 stipends to selected participants who would otherwise need to work over the summer. Note that its site has moved from summerscience.org to ssp.org, which is one reason older guides send you to dead pages.
What none of them publish
None of these four publishes an acceptance rate. You will find confident percentages in dozens of guides. Those numbers are estimates circulated between content sites, not figures the programmes release. Treat a programme as "very hard to get into" when the cohort is small and the applicant pool is national, and ignore the decimal places.
Type two: pre-college courses
These are paid summer courses run by university extension schools. They are legitimate, well-taught, and they are not selective in the way the first category is.
Harvard's Secondary School Program prices a seven-week residential place at $15,980 for 2027 — $7,960 in tuition for the required eight credits plus $7,820 for room and meals, with a $75 application fee and a $200 programme fee on top. The four-week residential option is $9,260. Commuting or online brings it down to between $4,180 and $8,160.
Brown's 2026 Pre-College fee schedule runs from $3,748 for one residential week of Summer@Brown to $10,858 for five weeks hybrid. Its Oxford-based BEE programme is $8,270 for twelve days.
These are real numbers from the universities' own fee pages, and they are worth looking at squarely. A fortnight at Brown costs more than a year of A-level tuition in most of the UK. That can still be a good use of money if you want the subject exposure and the independence. It is a poor use of money if you were buying selectivity, because you were not offered any.
Our breakdown of what research programmes actually cost sets these figures against the mentorship end of the market.
Type three: mentorship programmes
These pair you with a researcher — usually a PhD student or postdoc — to produce one piece of original work over several months. They cost less than pre-college courses and more than nothing, they are assessed rather than open-enrolment, and the output is a paper rather than a transcript line.
The honest summary is that the programme is a structure, not a credential. What a university reads is the work. Our comparison of mentored and self-directed research sets out when paying for that structure makes sense and when it does not.
Type four: the local placement nobody lists
The most under-used option is to email faculty at a university near you. It is free, it happens in your own city, and it does not appear in any ranking because nobody is marketing it.
Two constraints to know before you write. First, many wet labs cannot legally take you: the University of North Carolina's minors in laboratories policy sets out the age thresholds and supervision requirements that make under-18s difficult to host, and similar rules exist across US and UK institutions. Second, the reply rate is low. The University of Arizona's careers service recommends approaching three to five faculty members rather than one, which tells you what a normal outcome looks like.
Computational and archival groups face neither constraint as sharply, which is why they are the sensible place to start. Our guide to emailing a professor about research covers what to send.
How to choose
Work backwards from what you want to be able to say in twelve months.
If you want to be able to say I was selected, apply to the free selective programmes, apply early, and check the eligibility rules yourself rather than trusting a list. If you want to be able to say I did this and here is the paper, a mentorship programme or a local placement will get you there for less money. If you want subject exposure and a taste of university life, a pre-college course does that honestly — just do not expect it to function as a selection signal.
What does not work is the default: applying to whatever appears on the first page of search results in April, when the deadlines that mattered closed in December.
For the shortlist with dates and eligibility side by side, see our ranked list of the best summer programmes.
Frequently asked questions
What is the difference between a summer programme and a pre-college programme?
A pre-college programme is a paid summer course run by a university's continuing-education arm. You enrol, you pay, you take classes, and admission is mostly a question of meeting a modest academic bar. A selective summer research programme is a competitive placement in a working research group, usually free, where admission is the point. Both are marketed as summer programmes, which is why so many students apply to the wrong one.
Are free summer research programmes better than paid ones?
They are more selective, which is a different thing. A free programme like the Research Science Institute takes a small number of students and the selection itself carries weight. A paid programme can still produce good work if you do real research and finish something. What does not work is paying for a programme and assuming the name on the certificate does the arguing for you.
Can international students apply to US summer research programmes?
Usually not the free government- or university-funded ones. Stanford's SIMR restricts applications to US citizens and permanent residents living in the US, and the Clark Scholars Program at Texas Tech does the same with no exceptions. The Summer Science Program does admit international students. Check each programme's own eligibility page rather than an aggregator's summary.
When should I apply for summer programmes?
Earlier than most students expect. The Research Science Institute's 2027 applications close on 11 December 2026, around seven months before the programme runs. SIMR's 2026 deadline was 21 February. If you are thinking about next summer in March, you have already missed several of the most selective options.
Do summer programmes help with university applications?
The selective ones help because getting in is itself evidence. Paid ones help only to the extent that you produced something, which is why what you did during the programme matters more than which programme it was. UK personal statements and US supplements both ask what you contributed, not where you spent July.
What if I cannot get into a selective programme or afford a paid one?
Email faculty at your nearest university and ask about summer placements, or do independent research with public data. Neither needs an admissions committee's permission. The University of Arizona's careers service suggests contacting three to five faculty members rather than one, because the response rate on cold emails is low even when the request is reasonable.