Your supervisor has left, gone on long-term leave, or simply stopped replying — and the thesis has stopped with them. Nobody has told you what happens next, because the national frameworks do not cover it and your regulations are buried in a code of practice you have never opened. Here is what the rules actually say, and how to keep the thesis moving while the institution catches up.
Keep drafting while the handover happens — free tier, your scholarship, your structure held in one place.
What this costs if you wait for someone else to fix it
The costs are specific and they compound.
The first is time you will not get back. UKRI’s training grant terms build the entire extension architecture around the student’s own leave — family leave, medical leave, additional leave. There is no extension category for disrupted, absent or departed supervision. The responsibility for supervision sits with the research organisation under the grant conditions, which means a funded deadline continues to run while your supervision does not.
The second is drift. A doctorate is held together by a sequence of small decisions about scope, and those decisions stop being made when there is nobody to make them with. Three months of an unsupervised writing-up period produces a great deal of text and very little chapter.
The third is that the eventual replacement inherits a project they have to reconstruct from scratch, and reconstruction takes months you also do not have.
What your university is actually obliged to do
Every UK institution has rules for this. They differ so sharply that you cannot rely on anything you have heard from a friend at another university.
| Institution | Trigger for action on absence | What must happen |
|---|---|---|
| UCL | Principal supervisor absence “exceeding 4 weeks” | The supervisor must “Arrange (in consultation with the Subsidiary Supervisor and the Departmental Graduate Tutor) for a replacement”. A subsidiary supervisor must “Provide for continuity of supervision in the event of the absence or departure of the Principal Supervisor.” |
| University of Bristol | Unplanned absence “anticipated to exceed six months” | The Head of School “must consider whether to appoint a new permanent main supervisor or to explore other options with the PGR student.” Shorter absences trigger a duty to ensure “adequate supervision”, with “temporary supervisory support” if necessary. |
| Queen’s University Belfast | “a supervisor being unavailable for a significant period” — no numeric threshold | School responsibilities for “ensuring the appointment of appropriate supervision”, including “should a change in supervisor be required under different circumstances such as at the student’s request”. |
That first column is the finding. UCL imposes a duty to act at four weeks. Bristol imposes a duty merely to deliberate, and only once an absence is expected to run past six months. That is more than a six-fold difference in threshold, and a difference in kind — one is a self-executing obligation on your supervisor, the other is a consideration by a Head of School. Queen’s sets no number at all.
Bristol does state the underlying duty broadly, and it is the sentence to quote if you need one: the Head of School “is responsible for changing supervisors if this is necessary to ensure that a PGR student has continuous supervision during their registration.”
The clause nobody warns you about
One provision in Bristol’s code deserves isolating, because it can override everything above. On changing supervisors, the Head of School “must consider the requirements of any sponsors when changing supervisors. If a studentship is attached to a particular supervisor or to a specific research grant, it might not be permissible to change the student’s supervisors.”
If your funding came through your supervisor’s grant, the money and the supervision may be legally bound together. That does not leave you without options, but it does mean the conversation is different — and it is a conversation to have with your graduate school early rather than after the departure. If you are still weighing routes into a doctorate, the funding structures behind this are set out in UKRI studentship versus self-funded PhD.

The national frameworks are silent, and that is why this feels unmapped
It is not your imagination that there is no guidance. We looked for it.
The UKCGE Good Supervisory Practice Framework sets out ten criteria for doctoral supervision, from recruitment and selection through to reflecting upon and enhancing practice. None of them covers supervisor absence, departure, handover or continuity of supervision. The nearest it comes is an argument for team supervision: co-supervision developed to enhance the doctoral experience “by reducing their reliance upon a single individual and giving them access to a broader range of expertise and support”, with the honest caveat that it can also leave a candidate “playing ‘piggy in the middle’ to the detriment of their experience”.
The QAA doctoral degree characteristics statement — the edition we could open is the September 2015 second edition — likewise says that “The supervisor is fundamental to the support and development of the doctoral candidate” and promotes supervisory teams, but is silent on what happens when one leaves.
And PRES, the national survey of doctoral experience, does not measure it. PRES 2025 collected 35,475 responses across 93 institutions and reports supervision as its strongest area — 93 per cent agree their supervisors have the skills and knowledge to support their research, 92 per cent that they have regular contact. But it asks about contact and skills, never about change, absence or handover. A 92 per cent contact figure cannot tell you anything about what happens when contact stops. The wider picture from that survey is in our analysis of PhD wellbeing and supervision satisfaction.
So the protection is entirely institutional, it varies by a factor of six between neighbouring universities, and no national body has written it down. That is why you have to read your own code of practice, and why the practical burden of continuity falls further onto you than it should.
What to do, in order
- Find your own institution’s threshold. Open your research degree code of practice and search for change of supervisor, absence and continuity. You are looking for a number and for whose duty it is. Do this today, not when the departure is announced.
- Go to the graduate school, not the department. UCL gives candidates an explicit right “To consult in confidence with the Departmental Graduate Tutor if you are concerned about any behaviour of your supervisory team, or if you feel that a change of Supervisor is desirable for any reason.” Most institutions have an equivalent role. The formal escalation route, and how to use it well, is covered in managing your PhD supervisory team — this page picks up where a supervisor is gone rather than difficult.
- Ask about the funding clause explicitly. Whether your studentship is attached to a person or a grant determines what is possible. Ask in writing.
- Get the new arrangement recorded. Who is now the principal supervisor, who is subsidiary, what the meeting pattern is, and from what date.
- Check what it does to your registration. Interruption and extension are decided by your institution, not your funder, and the registration and continuation rules that govern them are set out in how long a UK PhD actually takes.
- Build the handover pack. See below. This is the part only you can do.
The handover pack that saves you a term
A replacement supervisor arrives knowing nothing about three years of work. Whether that costs you a fortnight or a term is decided almost entirely by what you hand them in the first meeting.
Keep it to a few pages:
- One page on the research question and the contribution, stated in a sentence each. If you cannot write these, that is the most useful thing the departure has told you.
- The chapter plan with a status against every chapter — drafted, redrafted, not started — and a word count.
- What was already agreed, with dates. Scope decisions, methods decisions, anything your previous supervisor signed off. Meeting records are worth their weight here.
- The data position. Collected, analysed, outstanding — plus any ethics approval and its expiry.
- A dated timeline to submission, so the conversation starts with a plan to react to rather than a blank page.

Keeping the thesis moving in the meantime
Here is the part that decides how much this actually costs you. The weeks between one supervisor leaving and another taking over are not dead time unless you let them be. They are, in fact, the ideal period for the work that does not need supervision: assembling the literature review into an argument, drafting the methods chapter from your own records, tightening chapters already written.
Draft your thesis in Tesify, concretely:
- Put the whole thesis in one workspace. The chapter plan, the sources and the drafts together — which is also, exactly, the handover pack described above. You build it once and it serves both purposes.
- Turn accumulated notes into chapters while there is nobody to review them, so that when a new supervisor arrives you are handing over drafts rather than intentions. The staged method for this is in three years of notes and no chapter.
- Let the bibliography format itself, so the months without supervision are not also months of accumulating clerical debt.
- Keep a dated record of what you drafted when. Useful for the new supervisor, and useful if you later need to evidence disruption to your graduate school.
The scholarship stays 100% yours throughout — which matters more than usual here, because with no supervisor reading your drafts the authorship question is one you must be able to answer entirely on your own account. Tesify is a drafting workspace, not a text generator; it holds the structure so that the thinking, which is the part nobody can do for you and the part your examiners will assess, is where your time goes.
There is a free tier, so you can build the handover pack and test the workflow on a single chapter before paying for anything.
Frequently asked questions
My supervisor is leaving. Is the university obliged to replace them?
Yes, though the wording varies. Bristol’s code makes the Head of School “responsible for changing supervisors if this is necessary to ensure that a PGR student has continuous supervision during their registration”. UCL places a duty on the subsidiary supervisor to “Provide for continuity of supervision in the event of the absence or departure of the Principal Supervisor.”
How long can a supervisor be absent before something must happen?
UK universities set wildly different thresholds. At UCL a principal supervisor must arrange a replacement for “absence for a period exceeding 4 weeks”. At Bristol an unplanned absence must be “anticipated to exceed six months” before the Head of School must even consider a permanent replacement. Queen’s University Belfast sets no number at all.
Will my funder extend my studentship because my supervisor left?
Almost certainly not. UKRI’s training grant terms build extensions around the student’s own leave — family, medical and additional leave — and contain no extension category for disrupted or absent supervision. Responsibility for supervision sits with the research organisation, not the funder.
Can I be refused a change of supervisor?
It is possible, and the reason is usually money rather than politics. Bristol’s code warns that the Head of School “must consider the requirements of any sponsors when changing supervisors. If a studentship is attached to a particular supervisor or to a specific research grant, it might not be permissible to change the student’s supervisors.”
Who do I speak to first?
Your graduate school or postgraduate research director, not your department head. UCL gives candidates an explicit right “To consult in confidence with the Departmental Graduate Tutor if you are concerned about any behaviour of your supervisory team, or if you feel that a change of Supervisor is desirable for any reason.”
What does the national guidance say about supervisor departure?
Remarkably little. The UKCGE Good Supervisory Practice Framework sets out ten criteria for supervision and none of them covers absence, departure, handover or continuity. The QAA doctoral characteristics statement is similarly silent. The rules that protect you are institutional, not national.
Why does team supervision exist?
Partly for exactly this situation. The UKCGE Framework records that the move to co-supervision came about to enhance the doctoral experience “by reducing their reliance upon a single individual”, though it also warns that co-supervision can leave a candidate “playing ‘piggy in the middle’”.
What should I put in a handover document?
Enough that a new supervisor can be useful within a fortnight: a one-page summary of the research question and contribution, the chapter plan with current status, what has been agreed and when, the data collection position, and a dated timeline to submission. Keep it to a few pages.
Does the PRES survey tell me how common this is?
No, and that gap is worth knowing about. PRES 2025 gathered 35,475 responses across 93 institutions and reports that 92 per cent of respondents agree they have regular contact with their supervisors. It does not ask about supervisor change, absence or handover at all.
Should I tell my examiners that my supervisor changed?
The thesis is examined on its own merits, and a change of supervisor is not a defence for weaknesses in it. What matters is that the work is coherent and yours. Raise the practical consequences with your graduate school instead, where extensions and interruptions are decided.
Is it academically legitimate to use an AI writing tool while I have no supervisor?
Provided it is a drafting workspace rather than a text generator, and provided you follow your institution’s rules on disclosure. The standard to hold yourself to is that you can explain exactly which work was yours — which is the same standard that applies with a supervisor in place.
Is my unpublished research safe in a writing tool, and what does it cost?
There is a free tier, so you can test it on one chapter before paying. On confidentiality, check any tool’s terms for what it retains and whether it trains on your text before you paste in unpublished findings or participant data, and anonymise transcripts first.


