PhD Research Proposal Guide: What Oxford, Cambridge, LSE, and ETH Actually Require

PhD Research Proposal Guide: What Oxford, Cambridge, LSE, and ETH Actually Require

A PhD research proposal must answer the instructions for your specific course or position. Oxford, Cambridge, LSE and European doctoral routes do not share one required format. Start with the actual document request, then show a worthwhile, feasible research project.

Our recommendation is to keep one underlying research plan and create a separate submission checklist for each target. Reusing your question can be sensible. Reusing an unchecked word limit, supervisor arrangement or combined-document format is not.

If the document names are unclear, first read research statement, SOP or proposal?. This guide covers the proposal and the application process around it, with official guidance checked on September 14, 2026. Oxford says its website is due to update for 2027–28 entry on September 16; recheck your live course page when that cycle opens.

Table of Contents

• How UK/EU PhD applications differ from US ones
• University of Oxford DPhil proposals
• University of Cambridge PhD proposals
• London School of Economics PhD proposals
• University College London PhD proposals
• ETH Zurich doctoral applications
• KU Leuven: the supervisor-first model
• How the proposal relates to your SOP
• Common mistakes to avoid

How UK/EU PhD applications differ from US ones

Country labels are an unreliable shortcut for the document you owe. A course recruiting students into existing laboratory projects can ask for research interests and preparation. A route built around an applicant's proposed project can require a detailed plan. You can encounter different arrangements within one university.

For Germany, begin with individual and structured doctoral routes. For any destination, identify whether you are applying to a programme, an advertised project or a prospective supervisor, and what happens after that initial stage.

Use this decision aid before drafting:

The table is an editorial aid. The current course or vacancy instructions control the actual requirement.

For an Australian comparison, the UQ EOI-to-application guide separates supervisor discussion, requested research material and the formal application rather than importing a UK proposal template.

University of Oxford DPhil proposals

Oxford's current research-proposal guidance explicitly says that not all research courses require a proposal. Some scientific courses instead ask about groups, supervisors, research interests or experience. The course page specifies the required document, length and assessment criteria.

Where a course requires both a personal statement or SOP and a proposal, Oxford's central instruction is to combine them with clear subheadings unless the course says otherwise. An unnecessary proposal can be removed; a substantially overlength proposal can make the application incomplete.

Do not use a university-wide word-limit list to override that course page. Record whether references, footnotes and other sections count. A proposal is also not a substitute for separately requested written work.

Structure and evaluation

Our recommended writing sequence is question, evidence gap, approach, feasibility and institutional fit. That is a planning sequence, not a compulsory Oxford template. Follow any structure the course actually specifies.

Explain why the research needs the department's expertise or resources, rather than replacing that explanation with institutional prestige. If you name an archive or facility, establish what you know about access and availability. A catalogue entry is not a promise that your proposed use is permitted.

Authorship and AI policy

Oxford's supporting-document policy permits AI tools for researching materials but says they may not “generate documents or form the basis of documents that are then adapted.” Follow that restriction and the requirement for your own work. The constructed cases below illustrate decisions, not submission text to adapt.

University of Cambridge PhD proposals

Cambridge's proposal guidance directs applicants to the course's requirements, including word or character limits and what counts toward them. It asks for an achievable project with an appropriate research approach and resources.

Use that guidance to test the connection between sections. Does the proposed evidence answer the question? Can the method produce the kind of claim you intend to make? What must happen before the main study can begin?

A project requiring access to an unavailable collection or specialist support that has not been identified has a feasibility problem. Writing a more confident introduction does not solve it. The data-access feasibility guide provides a detailed planning tool for that issue.

Cambridge also notes that topic changes after an offer may require Degree Committee approval. A proposal can develop through research, but do not promise yourself unlimited freedom to replace it after admission. If supervisory feedback changes the question substantially, resolve that change before treating different application versions as interchangeable.

London School of Economics PhD proposals

Read the instructions for the actual LSE department and route. LSE Management's guidance distinguishes its MPhil/PhD and MRes/PhD application materials. For the MRes/PhD in Management, it describes a proposal outline of no more than 1,500 words excluding the bibliography. That is not a university-wide LSE limit.

The department's proposal guidance addresses the research question, relevant literature, methods and contribution. It also explains that an initial proposal can evolve. That supports serious planning without pretending the application document is a completed thesis.

Our recommendation is to make the limitation in existing knowledge specific. “Little research exists” is weak unless you have established what the closest studies do and why they do not answer your question. Name the unresolved relationship, mechanism, comparison or interpretation, then explain the evidence needed to investigate it.

For methods you have not used, distinguish your current preparation from what you propose to learn. The unfamiliar-methods guide explains when a training plan is credible and when an entry requirement may make the project a poor fit.

University College London PhD proposals

UCL Psychology and Language Sciences' applicant guidance gives a specific example: its formal proposal is 1,000–1,500 words, including all sections and appendices, with key references. It asks for the question's importance and details of the studies and data, and describes initial contact with relevant potential supervisors.

This is divisional guidance, not a rule for every UCL doctorate. The page itself notes that research departments may have different requirements and directs applicants to check before submitting.

For a short proposal, preserve the design logic before adding a long literature review. A reader needs to understand what the first study would establish and how it leads to the next decision. If the progression only works when every initial result confirms your expectation, reconsider the design.

ETH Zurich doctoral applications

ETH's doctoral-position guidance describes recruitment by prospective supervisors and explains that a professor must be willing to supervise the dissertation. It directs applicants to relevant professorships and advertised positions; the Doctoral Administration does not arrange positions for applicants.

That does not mean no ETH application ever asks for a proposal. A particular position or doctoral programme can request research material. Read its current instructions rather than assuming that a generic project outline should be sent to central administration.

For an advertised position, our European PhD motivation-letter guide covers matching your preparation to the work. If a proposal is requested as well, establish whether you are developing an independent question or explaining an approach within an existing project.

KU Leuven: the supervisor-first model

KU Leuven's Arenberg Doctoral School applicant page describes applying online to a vacant Science, Engineering and Technology position or contacting a professor about opportunities. After selection or finding a supervisor, the departmental HR contact initiates the personal file and further instructions follow.

Keep that scope precise. It is an Arenberg process description, not a universal Belgian or European application procedure. Do not infer a standard four-year proposal template or document bundle from the institution's name.

The practical lesson is to identify the stage and contact responsible for the next action. A supervisor conversation, selection for a position and formal registration are related events, not interchangeable descriptions of admission.

How the proposal relates to your SOP

The proposal explains the research: what you want to investigate, why it matters and how the evidence could answer it. A personal statement or SOP usually has a different emphasis on your preparation and reasons for the application, subject to its actual prompt.

The same past project can appear in both documents without repeating the same paragraph. In the SOP, explain your contribution and development. In the proposal, explain the unresolved question that emerged and why the next study is needed.

Keep a submission matrix with the course, exact document name, limit, counted sections, upload arrangement and supervisor-contact instruction. Add the date you checked each source. Recheck it before submission, particularly when a new cycle has just opened.

Constructed case: one question, two different document packages

Imagine a fictional applicant developing a social-science question from a completed master's dissertation. One target asks for a proposal plus a statement in a combined file; another asks for a shorter proposal and a separate personal statement.

The applicant should not begin by shrinking the longer file until it fits both. They first identify the research argument that must survive: the question, the closest literature, the proposed evidence and the main feasibility dependency.

For the shorter proposal, they remove secondary background and a speculative extension while retaining enough method detail to make the central study assessable. In the personal statement, they explain their own contribution to the earlier dissertation and why the target's training matters. The combined-file version uses the required subheadings and respects each applicable instruction.

The project remains intellectually consistent across applications. What changes is the allocation of information and the verified institutional fit. If a university-specific resource is essential to one version, the other version needs its own feasible resource plan; changing the university name is insufficient.

Constructed case: an advertised project is not your independent proposal

Consider a second fictional applicant with a strong independent idea about environmental monitoring. They find an advertised position focused on a particular measurement problem and submit the independent proposal unchanged.

The document may be thoughtful yet fail to answer the vacancy. The applicant needs to decide whether the advertised work actually interests them and whether their preparation fits its requirements. If it does, the requested letter or research statement should explain that connection and address the project's scope.

If the position invites an original proposal, they can develop one within the stated remit. If it does not, attaching an unrelated plan does not demonstrate initiative automatically. It can leave the reader unable to assess the applicant for the work being recruited.

This case requires a choice of research direction before document adaptation. It is different from the first case, where the same underlying question could fit two verified application formats.

Common mistakes to avoid

The most damaging problem is a broken connection between the question and the evidence. An ambitious title cannot compensate for a design that studies something else. Read each section as a claim about what the project could establish, then check whether the next section supplies the necessary support.

A second problem is unearned certainty. Distinguish confirmed access from an enquiry, completed training from planned learning and your own work from a team's output. Describe unresolved dependencies and meaningful alternatives rather than promising that every approval or resource will arrive on schedule.

A third is treating feedback as a request for cosmetic editing. If a supervisor says the project is too broad or the data cannot answer the question, a shorter introduction is insufficient. Use the supervisor-feedback revision guide to connect each comment to a research decision.

Finally, avoid copying samples. A proposal example can show how a question and method relate; it cannot supply your evidence, research judgement or institutional permissions. Check quotations and citations yourself, and do not share confidential or restricted material in an application or review upload without authorization.

How GradPilot can help

GradPilot's standalone PhD research-proposal review examines the writing around the research problem, literature positioning, design, feasibility and fit. It does not validate a scientific method, grant data access or decide admission. Follow the target's authorship and assistance rules, including Oxford's published restriction above.

Use the PhD review page to choose the actual document you need reviewed, and the PhD application hub for related preparation. For previous research narratives, see PhD SOP examples and analysis; for a different document request, return to the SOP and personal-statement comparison.

A useful proposal gives the reader a worthwhile question, a defensible plan and an accurate account of what still needs to be resolved. The appropriate length and packaging come from the course or position you are applying to.