How to Write a Research Statement for a PhD Application (2026)

This is the full formal document submitted with an application file, not the three-to-five sentence paragraph that belongs inside a cold email. If you're looking for that shorter version, our guide to the cold-email research statement covers it directly. This post is about the standalone document itself: how long it actually needs to be, how that length changes by country and field, what structure reviewers expect, and how to write one honestly when you don't yet have a fully formed dissertation topic.

First, work out which document your program is actually asking for

A meaningful share of confusion around this document comes from programs using different names for overlapping but distinct things. A statement of purpose is generally the more narrative document, covering your academic background and why a specific program fits your goals. A research statement is expected to argue an actual research direction, sometimes with methodology, sometimes with a specific proposed project. Some programs ask for one document that does both jobs. Others split them explicitly: the University of Chicago's Division of the Social Sciences asks applicants for a candidate statement that must include a concise, well-articulated research proposal as a distinct component within it, capped at 750 words total. There is no single universal template here, and the single most useful thing you can do before writing anything is read the specific program's own instructions for the exact document name, word limit, and required content, rather than assuming a format from a different program applies.

How long this actually needs to be, by country

Length expectations differ enough between the US and UK systems that using the wrong country's norm as your default is a real, common mistake. In the US, most programs land somewhere between 500 and 1,000 words. MIT's statement of objectives runs around 1,000 words. The University of Chicago's research proposal component, nested inside its broader candidate statement, is capped at 750 words total for the whole statement. Field convention shifts this too: humanities programs often expect something closer to one to two pages, social science programs commonly run two to three pages, and STEM or lab-science programs often run three to four pages, frequently broken into subheadings for separate project lines, background, and proposed methods.

In the UK, and across much of Europe, the research proposal isn't a supporting document, it's the centerpiece of the application, because UK PhDs are typically structured around a specific, supervisor-approved project rather than a general statement of academic interest. That shows up in the expected length: the University of York asks for 1,000-1,500 words, with a hard maximum of 1,500 including references. Oxford's School of Geography and the Environment suggests around 2,500 words with no fixed upper or lower limit. Cambridge's requirement varies meaningfully by faculty, an 800-word proposal for the Faculty of English versus 1,500 words for the Faculty of Education. If you're applying to a UK program, treat the specific department's stated word count as the real target, not a generic "1-2 pages" rule that describes the US system instead.

A structure that holds up across most versions of this document

Regardless of the exact length your program expects, most versions of this document do the same four jobs, roughly in this order:

1. The specific question or gap. Open with the actual research question or problem you want to work on, stated concretely enough that a reviewer in the field could evaluate whether it's interesting and unsolved, not a broad statement about the field's general importance.

2. Context: why this question, why now. A short section situating your question against what's already known, naming specific work or a specific gap it leaves open, not a full literature review. This is where citing genuinely recent work in the area matters more than citing the most famous work, since it signals you're tracking the current state of the field rather than what was taught in an undergraduate course.

3. Your proposed approach or methodology. How you'd actually go about answering the question, at whatever level of specificity your discipline expects. STEM and lab-science statements typically need real methodological detail here, proposed experiments, datasets, or models. Humanities and qualitative social science statements can stay more conceptual, focused on theoretical framework and sources, as long as the approach is genuinely thought through rather than vague.

4. Fit: why this program, and where relevant, why this supervisor. A specific connection to resources, faculty, or infrastructure at the target program, not a generic sentence about the university's reputation. In the UK model, this section often needs to name a specific proposed supervisor directly, since the application itself may be evaluated by that person.

The mistake that shows up most in STEM statements: an overloaded methods section

A common failure mode in technical fields is treating the research statement like a grant proposal aimed at funding agency reviewers who already share your background knowledge, packing in every technical detail of a proposed method before establishing why the underlying question matters. A PhD admissions committee is reading dozens or hundreds of these, and a methods section that opens with acronyms and technique names before any accessible framing of the actual research question loses the reader before the interesting part. Lead with the question in plain terms, then get technical once the reader understands why the technical detail matters.

The mistake that shows up most in UK proposals: a project too ambitious to actually finish

Because a UK proposal is being evaluated as a genuinely proposed, fundable, doable project rather than a general statement of interest, the most common weakness is scope: a research question broad enough to describe an entire subfield's worth of open questions rather than a project one person could realistically complete within a funded PhD's typical three-to-four-year timeline. A UK supervisor reading a proposal is assessing feasibility as directly as they're assessing originality, so naming a clearly bounded question, with an honest sense of what a completed dissertation from this project would actually look like, reads as more credible than an ambitious-sounding proposal with no realistic endpoint.

Should you name a specific supervisor in the document itself?

In the UK model, generally yes, since the application is often reviewed by the named supervisor directly and the proposal is expected to connect to their actual expertise or lab resources. In the US model, this varies more by program: some explicitly want you to name specific faculty whose work overlaps with your interests, while others discourage it in favor of demonstrating fit with the department broadly. Either way, if you do name a specific person, the reference needs to point at something real and current in their work, the same standard that applies to a cold email's research paragraph, rather than a generic mention of their general research area.

Writing this honestly without a finished dissertation idea

No admissions committee expects an incoming PhD student to already have a finished, defensible dissertation plan, and pretending otherwise usually backfires. The honest version of this document names a genuine open question or methodological gap that draws you in, backed by real evidence you've engaged with the current literature, and is upfront that the specific shape of the eventual project will develop in conversation with an advisor. That reads as realistic research judgment, not underpreparation. What doesn't work is inventing an overly specific, fully mapped-out project you haven't actually thought through in order to sound more prepared, since a knowledgeable reviewer will often notice the gap under a single follow-up question during an interview.

How this connects back to your outreach

If you've already been cold-emailing professors before applying, the research direction in your formal statement should genuinely connect to those conversations, not restate a completely different interest out of nowhere. The 3-5 sentence research paragraph used in a cold email and the full formal statement should feel like they came from the same person thinking about the same question at different levels of depth, not two unrelated documents written for two different audiences. If your CV also needs rebuilding for the formal application file, our academic CV template for a PhD application covers the full-length version, distinct from the shorter CV built specifically for a cold email.

Getting to the point of writing a real, specific research statement usually starts earlier, with finding professors whose current work actually overlaps with a direction you can defend. GradScoutFunding searches professors by field and country, confirms they're actively publishing, and drafts a personalized first email grounded in their real recent paper for you to review and send yourself, it never sends anything automatically. The free tier gives 2 credits with no card required, and paid credit packs are one-time purchases that never expire.

Common questions