Write to postdocs and early-career lecturers rather than professors, at universities outside the top handful, about a specific paper you have actually read. Expect most emails to go unanswered — the reply rate on a well-written cold email is low, and the students who succeed are the ones who send twenty rather than three. The whole thing costs nothing but nerve.
Every year, a small number of school students end up working with a real academic on real research, unpaid, because they sent an email.
Almost nobody tries. The ones who do mostly send the wrong email to the wrong person, get no reply, and conclude it does not work.
It does work. It just has a low hit rate, and the mechanics matter more than most students expect.
Who should you actually write to?
Not the professor whose book you have heard of.
Senior academics at famous institutions receive a great deal of unsolicited email and have the least discretionary time. A student writing to the head of a well-known lab at Oxford is joining a queue they cannot see.
Write instead to:
Postdoctoral researchers. They are running the actual experiments, they have published recently, they are often building supervision experience deliberately, and they receive almost no cold email. This is the single highest-yield group and almost nobody targets it.
Early-career lecturers. Recently appointed, building a group, and often keen on outreach because it strengthens grant applications.
Academics at universities outside the top ten. A researcher at a solid civic university does work of genuine quality and gets a fraction of the inbound mail a Russell Group professor does. Your email stands out in proportion to how quiet their inbox is.
People whose recent paper you can actually read. This is the binding constraint. If you cannot get past the abstract, you cannot write a credible email about it.
How do you find their recent work?
Three routes, none of which requires a subscription.
Google Scholar. Search your topic, sort by recent, and read abstracts until something interests you. The author list gives you names; the university page gives you the email address.
arXiv, bioRxiv, medRxiv, SSRN. Preprint servers, free and complete, covering physics, maths, computer science, biology, medicine and the social sciences. Everything is readable in full.
Departmental pages. Every UK university lists its academics with research interests and recent publications. Pick a department, read twenty profiles, shortlist five.
Then read the paper. Not the title — the paper, or at minimum the abstract, the figures and the conclusion. The email you write afterwards will be obviously different from one written by someone who read a title.
What goes in the email?
Short, specific, and asking for something small.
The structure that works:
Subject line: specific and plain. "Year 12 student — question about your 2026 paper on [topic]" beats "Research opportunity request."
First sentence: who you are, in one line. Year group, school, subject. No preamble about your lifelong passion.
Second: the specific thing you read and one genuine thought about it. This is the whole email. It is the only part that distinguishes you from a mass mailing, and it cannot be faked without doing the reading.
Third: what you are asking for, made small. Not "will you supervise my research project." Something like: could I ask you two questions about the method, or is there anything a Year 12 could usefully help with over the summer.
Fourth: what you can offer, honestly. Hours a week, relevant skills, whether you can code or handle data. Be accurate — overstating this is the fastest way to waste everyone's time.
Close: thank them, say you understand if they are too busy, and stop. Total length: under 200 words.
A worked example
Constructed to show the shape, not a real exchange.
Subject: Year 12 student — question about your paper on urban heat islands
Dear Dr [Name],
I am a Year 12 student at [School] studying Geography, Biology and Maths. I read your 2026 paper on night-time temperature retention in mid-sized UK cities.
I was surprised that the effect held after you controlled for building density — I had assumed density was doing most of the work. If I have understood the discussion correctly, you suggest surface material matters more than volume, which was not what I expected.
I am hoping to do an independent research project this year on something related. Could I ask you two short questions about how you selected the comparison cities? And if there is any small task a Year 12 could usefully help with over the summer — data cleaning, literature searching — I would be glad to.
I have about six hours a week and some experience with Python and spreadsheets. I completely understand if you are too busy to reply.
Thank you for your time,
[Name]
Notice what is absent. No mention of university applications. No flattery. No CV attached. No request for a reference.
What gets an email deleted?
Six things, and most students do at least three of them.
"Dear Professor" with no name, or the wrong name, or the wrong title. Check whether they are a Dr or a Professor; getting it wrong signals a mail merge.
Anything that reads as mass-sent. No specific paper, no specific idea — an academic can spot this in four seconds.
Asking for too much. "Would you supervise my research project" is a large commitment to a stranger. "Could I ask two questions" is not.
Length. Over 300 words and the reply gets postponed, then forgotten.
Mentioning your university application. True, understandable, and it reframes you from interested person to someone seeking a credential.
Attaching a CV. You are not applying for a job. It signals a transaction.
What reply rate should you expect?
Low, and you should plan for that rather than be discouraged by it.
We are not aware of any reliable published data on cold-email response rates for school students, and anyone quoting you a precise figure is guessing. What is safe to say is that most emails go unanswered, that a minority get a courteous no, and that a small number get a real conversation — and that the variable you control is volume and quality, not luck.
The practical implication: send twenty, not three. Students who send three and stop conclude the approach does not work. Twenty personalised emails is perhaps six hours of work spread over a fortnight, and it only has to work once.
Keep a simple spreadsheet: name, institution, paper, date sent, reply. Follow up once after two weeks, briefly, and then let it go.
What to do with a reply
Most replies will be short, and that is normal.
If they answer your questions but offer nothing else: thank them, and ask one good follow-up. You have started a correspondence, which is worth more than it looks. Several students have turned a two-email exchange into a summer placement the following year.
If they offer a small task: do it well and on time. That is the entire audition.
If they say no: ask whether they can suggest a colleague or a group that takes school students. People who cannot help are often glad to redirect.
If they offer something too big: say so honestly. Agreeing to work you cannot deliver ends badly for both of you.
What to do if nobody replies
You have lost a fortnight and gained a genuinely useful skill, and there are other routes.
Structured free programmes. UK students have a real free tier — UNIQ, Sutton Trust summer schools, In2STEM and research placements. We set out who qualifies for each in free research opportunities for UK students.
Your own school. An EPQ is a supervised independent research project your school may already run, and it carries UCAS points. See how to turn your EPQ into a publishable paper.
Your school's alumni network. Frequently the single most underused resource in a sixth form. Someone three years above you is now a PhD student somewhere, and will reply.
Paid mentorship. We run the Oxford Scholar Programme for students with the appetite but no access to a research environment. It is the most expensive option on this page and we would rather you tried the others first — the honest comparison of what everything costs is here.
One thing worth saying plainly
A cold email that works is a better story than anything you can buy.
It demonstrates initiative, which is the one quality no programme can sell you, and it is the kind of specific detail that makes a UCAS Question 3 answer credible. "I emailed twenty researchers, one replied, and I spent the summer cleaning her dataset" is a paragraph an admissions tutor remembers.
Send the emails first. Pay for help only if they do not work.
Frequently asked questions
Will a professor really reply to a school student?
Some will. The odds improve sharply if you write to a postdoc or early-career lecturer rather than a senior professor, choose someone outside the most-contacted institutions, and reference a specific recent paper you have genuinely read. Most will not reply, which is why volume matters.
What should the subject line say?
Something specific and plain — your year group and the paper you are writing about. Avoid anything that reads like a template, and never write "Research Opportunity" as a subject line.
Should I attach my CV or transcript?
No. It makes the email read like a job application and lengthens something that should be scannable in thirty seconds. If they want to know more, they will ask.
How long should I wait before following up?
About two weeks, once, in three lines. If there is no reply to the follow-up, move on — academics are busy rather than rude, and a second chase does not help.
Is it better to email someone at a famous university?
Usually not. Researchers at less-contacted institutions do excellent work and have far more capacity to answer. The quality of the science is not proportional to how often you have heard of the university.
What if I do not understand the paper well enough to write about it?
Read the abstract, the figures and the conclusion, and pick a paper where that is enough to form one honest thought. If you cannot get that far with anything in the field, you are looking at work too advanced for a first project — try a different subfield rather than bluffing.