Research opportunities for high school students: every route, ranked

Most guides to this topic are lists of programme names with the deadlines copied off each website. That is not useful, because the names are the easy part. What decides whether you end up doing research is a set of constraints nobody puts on the page: citizenship, geography, grade level, and whether the programme has any obligation to tell you how selective it is.

So this is organised by route rather than by name, and every figure is sourced.

What research opportunities actually exist for high school students?

Four routes, and they differ more in who they exclude than in what they produce. Free university-run programmes are the most prestigious and the most restrictive. Paid mentorship programmes are the most accessible and the most variable in quality. Local lab work is the cheapest and the hardest to secure. Independent projects have no gatekeeper at all.

The rest of this page works through each.

How competitive are the free university programmes?

More competitive than almost any undergraduate admissions process, and mostly unmeasurable. Of thirteen major programmes we checked, only two publish an acceptance rate. The Research Science Institute reports roughly 2,500 US and 600 international applications for about 100 places — under 2.5%. Stony Brook's Simons Summer Research Program states approximately 5%.

The other eleven publish cohort sizes and nothing else. The Clark Scholars Program at Texas Tech takes twelve students. Stanford's SIMR takes around fifty. MIT PRIMES says only "a small number." You can infer that these are extremely selective, but you cannot calculate anything, and neither can the blogs that quote percentages for them.

This matters practically. If you are building a shortlist, you are choosing between programmes whose odds you genuinely cannot compare.

Three things are worth knowing before you apply anywhere:

Citizenship rules exclude most of the world. MITES, MOSTEC, Stanford SIMR, the Clark Scholars Program and the Simons programme all require US citizenship or permanent residency. MIT PRIMES requires US residency. RSI takes international students only through partner-country nominating bodies, not direct application.

Several "free" programmes are not free. Stony Brook's Garcia Summer Program charges a $4,000 lab fee plus application and materials fees, with housing at $2,599 on top. UT Austin's High School Research Academy costs $4,000. MIT Beaver Works is free only for families under $200,000 in income and $3,000 above it. Each of these is routinely described as free in programme roundups.

NASA OSTEM does not take high school students. NASA's own eligibility pages require enrolment at an accredited college or university. A large number of blog posts and listicles still describe OSTEM as a high school internship. They are wrong, and applying costs you a deadline you could have spent elsewhere.

What do paid research mentorship programmes cost?

Between roughly $3,000 and $11,800 for the ones that publish a price, and several do not publish one at all.

Pricing opacity is the norm in this sector, not the exception. Four of the seven above will not tell you what their programme costs until you speak to someone.

The industry also has a documented credibility problem. In May 2023, ProPublica and the Chronicle of Higher Education published an investigation by Daniel Golden and Kunal Purohit into paid research programmes, finding roughly twenty companies serving around 12,000 students a year at $2,500 to $10,000 each, with mentors often doctoral students paid $150 to $200 an hour. Its sharpest finding concerned publication: the Journal of Student Research accepted 65% of submissions from one programme's students, with edits requested "nearly never."

That investigation is now three years old, and no comparable follow-up has been published — searches for one return almost entirely marketing content from the companies themselves. Treat its specific figures as dated and its structural criticism as live.

One programme responds to this directly. Pioneer Academics states plainly that it does not offer "pay-to-play" publication opportunities and declines to push students toward journals at all. Whether that makes it better depends on what you want, but it is the clearest position anyone in the sector takes.

Can you just email a professor and work in their lab?

You can, and a small number of students succeed, but the realistic response rate is low and the obstacle is usually institutional rather than personal.

The most honest account comes from the Goldhaber-Gordon lab at Stanford, which publishes its own policy: it receives "tens of such requests per year," takes a high school intern "only once every 4 to 5 years," and tells students "don't be surprised or offended if you don't get a reply." No university publishes a measured response rate, but that ratio is the most specific number any lab has put in public.

The barrier students rarely anticipate is safety policy. UNC's rules for minors in laboratories prohibit work involving BSL-2 agents, radioactive materials or animals, require constant direct supervision, and require environmental health paperwork filed a month before the student starts. A willing professor frequently cannot take you regardless of willingness.

If you try this route anyway — and it is worth trying, because it costs nothing — the University of Arizona's guidance suggests contacting three to five faculty members rather than one. Read a paper of theirs first, and say something specific about it. A message that could have been sent to any academic gets treated like one.

Which route should you choose?

Apply to the free programmes you are eligible for, because they cost nothing but time, and assume you will not get in. Then choose a fallback deliberately rather than by default.

Choose a local lab if you live near a research university, have a specific field in mind and eighteen months of runway. The experience is real and the cost is zero. Accept that you may spend the summer doing useful but unglamorous work, and that authorship is unusual.

Choose an independent project if you are self-directed and your field does not require equipment. This is the route most under-recommended relative to how well it works, particularly in economics, computer science and the social sciences.

Choose a paid programme if you need structure, a deadline and someone who will read your drafts — and only after you have established that the free routes are closed to you. Ask what you will produce, who exactly will supervise you, and what happens if the work is not good enough to submit anywhere. Programmes that will not answer the third question in writing are telling you something.

Where OCAR fits

The Oxford Centre for Advanced Research runs 1:1 mentorships pairing students with PhD candidates and early-career researchers at Oxford, Cambridge, other leading UK universities and Ivy League institutions. Two programmes: the Oxford Scholar Programme at £2,000 (roughly $2,600) with ten contact hours across ten to fourteen weeks, and the Oxford Publication Fellowship at £3,400 (roughly $4,400), which supports submission to a peer-reviewed student journal.

Our acceptance rate is approximately 60%, self-reported and unaudited, exactly like every other figure in the table above. We publish it because a programme that hides its selectivity while selling you selectivity is worth less than one that does not.

Two honest limits. If RSI, PRIMES or SIMR are open to you, they are better than what we offer and they are free — apply there first. And if you want guaranteed publication, no reputable programme can offer it, ours included; we submit to the Oxford Journal of Student Scholarship, the Journal of Research High School and Vanderbilt's Young Scientist Journal through their standard review processes, which reject student work regularly.

What to do this month

If you are a junior in the US, the free-programme deadlines are the immediate constraint: MIT PRIMES closes on 2 November 2026, and most others open between December and February. Those are the applications that cannot be made later.

Everything else can wait a fortnight while you work out which field you actually want to spend three months on. That question is harder than it sounds, and getting it wrong is the most common reason students abandon a project halfway.

Frequently asked questions

Do colleges actually care about high school research?

Some do, explicitly. Caltech reports that 42% of admitted students submitted research, but its guidance is specific: the work must come from 'a laboratory in a university or research-affiliated organization' and be accompanied by a letter from the research site. Independent projects go in a portfolio instead. Most universities publish nothing this precise, so treat confident claims about how research is weighted with suspicion.

Is it too late to start research in eleventh grade?

No, but the calendar tightens. Most free university programmes require you to be a junior when you apply, meaning the application window is the winter of eleventh grade for the following summer. A mentored project of ten to fourteen weeks can still be completed in the spring or summer of eleventh grade and written about in an application submitted that autumn.

How competitive are the free programmes really?

The two that publish figures are brutal. RSI receives roughly 2,500 US and 600 international applications for 100 places, an acceptance rate below 2.5%. The Simons Summer Research Program states approximately 5%. The other eleven programmes we checked publish only cohort sizes, so any percentage you see quoted for them comes from a blog rather than the programme.

Can international students access US research programmes?

Rarely, and this is the constraint most guides omit. MITES, MOSTEC, Stanford SIMR, Clark Scholars and the Simons programme all require US citizenship or permanent residency. RSI admits international students through partner-country nominating bodies rather than direct application. The Summer Science Program and the Garcia programme are the notable exceptions that accept international applicants directly.

Do I need a lab to do real research?

Not for every field. Economics, psychology, public health, political science, linguistics and computer science all support projects built on public datasets, survey work or computational methods. Wet-lab science is the field where the access problem is genuine, and it is also the field where university minor-in-lab policies are most restrictive.

What does a research project actually produce?

Usually a paper of 4,000 to 8,000 words, sometimes a poster or a conference presentation. Whether it can be submitted to a journal depends on the work: original data or a genuinely novel analysis can be; a literature review generally cannot, because most student research journals want new findings rather than summaries of existing ones.