Gates Cambridge Research Proposal Guide (PhD)

Finding a Supervisor at Cambridge

Start at least one month before the deadline

Identifying and contacting a potential supervisor isn’t something you can do the week before applications close. Cambridge academics are busy, they receive many enquiries, and building a genuine connection takes time. Start browsing faculty profiles and reaching out at least one month before your deadline — ideally two or three months. For 2027/28 entry, US-round applicants (deadline 14 October 2026) should already be in touch with supervisors; international-round applicants should reach out by early November or early December 2026 at the latest (deadlines 8 December 2026 or 6 January 2027).

A supervisor who is enthusiastic about your project dramatically strengthens your application. It tells the department — the first filter — that a real person at Cambridge wants to work with you. Without supervisor buy-in, even a brilliant proposal can fail to secure a department nomination.

How to identify potential supervisors

• Browse the Cambridge Research Portal for faculty in your area.
• Read recent publications by Cambridge academics in your field.
• Check departmental websites for research group pages.
• Attend online seminars or conferences where Cambridge faculty present.
• Ask your current professors if they have Cambridge contacts.
• Look at recent PhD theses from the department for advisor names.

The initial email

Your email to a potential supervisor should be concise (under 300 words), specific, and show that you’ve read their work.

• A one-sentence introduction of yourself and your background
• A reference to their specific work that interests you (name the paper)
• A brief description of your proposed research (3–4 sentences)
• A clear question: “Would you be open to discussing potential supervision?”
• Don’t attach your full CV unsolicited — wait until they ask
• Don’t send identical emails to multiple supervisors in the same department

Writing the Research Proposal

What “original research” actually means

Your proposal must show that you’re proposing original research — work that will contribute new knowledge to your field. That’s different from assisting with someone else’s project. “Lab assistance is not original research,” as one department admissions officer put it bluntly. You need a research question that is genuinely yours, even if it builds on or extends your supervisor’s work.

Originality can come from many places: a new method applied to an existing question, an existing method applied to new data, an entirely new question nobody has asked, or a novel synthesis of existing approaches. What matters is that you can say what is new about your contribution.

Proposal structure

Format varies by field, but most strong proposals include these elements:

1. 1
   Research question

   A clear, focused question that can be investigated within 3–4 years. Not too broad (“What causes climate change?”) or too narrow (“Does this specific gene variant affect this specific pathway in this specific organism?”).

2. 2
   Literature review

   Show awareness of existing work and identify the gap your research fills. This demonstrates you’re not reinventing the wheel.

3. 3
   Methodology

   How will you answer your question? What data, tools, frameworks or approaches will you use? Be specific enough to show feasibility.

4. 4
   Timeline

   A realistic schedule for a 3–4 year PhD, with milestones: literature review, data collection, analysis, writing.

5. 5
   Significance

   Why does this research matter? Who benefits? This connects directly to the “improving lives” Gates criterion.

6. 6
   Cambridge-specific resources

   What at Cambridge makes this research possible? Name labs, datasets, supervisors, centres or archives.

Hypothesis-driven vs. exploratory

The strongest proposals are hypothesis-driven: they state a clear prediction or argument and describe how the research will test it. Exploratory proposals (“I want to investigate the relationship between X and Y”) are weaker because they don’t show analytical clarity.

That said, not every field needs a hypothesis in the traditional sense. In the humanities, a clear interpretive framework or argument does the same job. In computer science, a well-defined system architecture or algorithmic contribution works. The point is to show you have a direction, not just a topic.

Common Pitfalls

• Too ambitious. Proposing to solve a 50-year-old problem in 3 years signals that you don’t understand the scope of the challenge. Be ambitious but realistic: better to investigate one well-defined question deeply than to address three superficially.
• Too generic. If your proposal could go to any university unchanged, it’s not Cambridge-specific enough. Name the people, labs, datasets and resources at Cambridge that make your research feasible.
• No methodology. An interesting question without a plan to answer it is the most common weakness. Show that you know how to do research, not just what to research.
• Ignoring limitations. Every project has limitations. Acknowledging them shows intellectual maturity; ignoring them makes you look naive or inexperienced.
• Jargon overload. Gates selectors include people outside your field. Make your proposal accessible to an educated non-specialist. Technical precision is important; impenetrable jargon is not.
• No supervisor mentioned. Not naming a potential supervisor suggests you haven’t engaged with the Cambridge community. Even without formal agreement, mention faculty whose work aligns with yours.

Your research proposal is crucial, but it’s only one part of the application. Make sure your essays and references are equally strong.