Can I reuse my own paper?
Usually not without asking first, and the reason is not the one people assume. Nobody thinks you stole from yourself. An assignment is a measurement of what you learned in that module, and work that has already earned credit cannot measure it a second time, so nearly every university treats resubmission as an integrity matter with a name of its own. Permission is granted more often than students expect. It has to be granted in advance, though, and Turnitin will notice either way: your own earlier submission is stored, and across two different classes it is not automatically excluded from your similarity report.
The question usually arrives in one of three shapes. You wrote something last year that fits this term's brief. You are taking two modules whose assignments overlap. Or you are retaking a module and you still have the essay. All three have the same answer and the same fix, which takes about two minutes.
The word you need to search for
Start with a practical problem: your university almost certainly has a rule about this, and searching your handbook for the phrase you have in your head will probably miss it. The offence goes under at least four names.
Self-plagiarism is the common one in the United States and Australia. Auto-plagiarism is what Oxford calls it. Multiple submission is the formal term in a lot of American codes, sitting in a numbered list of definitions next to fabrication and unauthorised collaboration. Duplicate submission and recycling both turn up too. If a search for one returns nothing, try the others before concluding your institution has no policy, because the odds are it does.
The naming matters for a second reason. A rule filed under plagiarism sounds like theft and reads as an accusation. A rule filed under multiple submission is closer to what is actually going on, which is an administrative constraint on how credit is awarded.
What nine universities actually say
We read the published policies at nine universities and pulled the operative sentence from each. They are quoted exactly, and every one links to the page it came from.
| Institution | What the policy says | Permission needed |
|---|---|---|
| Oxford | "You must not submit work for assessment that you have already submitted (partially or in full), either for your current course or for another qualification of this, or any other, university." | Only where course regulations allow it |
| Cornell | Work used to determine a grade "may not be submitted by that student in a second course, unless such submission is approved in advance by the faculty member in the second course." | In advance, second instructor |
| Villanova | "Students shall not submit the same academic work, or substantially the same work, for more than one class course without prior approval from both faculty members." | In advance, both instructors |
| Penn State | "Students may not typically use the same work, or part of the same work, for credit in multiple classes." Named examples include retitling an old essay and pasting from a previous assignment. | In advance, current instructor |
| Buffalo | "Re-submitting academic material (in whole or in part) that has been previously submitted by the same student without prior and expressed consent of the instructor." Applies within a single course too, and when retaking one. | In advance, always |
| South Florida | "Multiple submissions shall include the use of any prior academic effort previously submitted for academic credit at this or a different institution." | Prior written approval |
| Towson | "Submitting substantial portions of the same academic work (including oral reports) for credit more than once without authorization." Adds: "Building upon or reworking prior work is acceptable with permission of the instructor." | In advance, and reworking is explicitly allowed |
| Northeastern | "Handing in the same academic material, in part or in total, for more than one course without explicit permission from the instructor(s)." | Explicit permission |
| UNSW | "Re-using all, or parts, of a body of work that has already been submitted for assessment without proper citation." | Turns on citation and disclosure |
Nine for nine, reuse without asking is a breach. That is a rare level of agreement: compare it with something like how many sources a paper needs, where the same nine institutions would give you nine incompatible numbers. Here the rule is genuinely settled, and the settled part is the timing. Permission counts when it arrives before the submission.
Villanova is the only one in the set that explains itself, and the explanation is the whole rationale in two sentences: "Faculty create assignments in order to foster a particular kind of learning in a course. Handing in work done for a previous course may preclude this learning." Nothing there is about ownership. It is about what the assignment is for.
The three edges where policies disagree
The core rule is uniform. Three specific situations are not, and they are exactly the situations students find themselves in.
Reuse inside one course. Most policies address two different courses. Buffalo's covers reuse "either in a single course or in multiple courses", so recycling your own week-three response paper into your week-nine essay is in scope there and might not be elsewhere. If your two assignments are in the same module, do not assume the rule is only about crossing module boundaries.
Work from another institution. Oxford and South Florida both reach across institutions explicitly, South Florida naming "academic credit at this or a different institution" and Oxford naming "another qualification of this, or any other, university". If you transferred, or you are doing a masters after a bachelors elsewhere, your old essays travel with you as far as the policy is concerned.
How much counts as the same. "Substantially the same work" (Villanova), "substantial portions" (Towson) and "in part or in total" (Northeastern) are all doing the same job, which is to close the paraphrasing loophole before anyone tries it. There is no percentage anywhere in these documents, and Towson says openly that what counts as a substantial portion "is determined solely by the university". Rewriting the sentences does not convert a reused paper into a new one.
The one edge everybody gets wrong: retaking a module. It feels like the same course, so reusing your own essay feels like resubmitting a draft. Buffalo addresses this directly, requiring students to ask "even when retaking a course for a second time", and the new instructor is the person who has to agree.
Turnitin matches you against yourself
The policy is one half. The mechanics are the half that produces the surprise, and Turnitin's own documentation is clear about them.
Inside one class, you are protected. Turnitin applies what it calls a self-match exclusion by default, "designed to allow students to submit early drafts of their work without their final submissions matching against their own previous drafts". Those matches still exist. They are hidden in the Excluded sources section of the similarity report, and an instructor can put them back.
Across classes, you are not. In Turnitin's words, an automatic exclusion "does not occur if a student submits the same paper to assignments across different classes, provided at least one of those assignments was configured to save papers to the standard Turnitin repository". The standard repository is the global one: Turnitin's assignment settings describe it as storing submissions "in the global Turnitin repository", as opposed to an institution-only repository or not storing them at all. Most assignments use it, because that is the setting that makes cross-institution matching work in the first place.
Which produces the situation this article exists for. You submit an essay you wrote two years ago, at this university or another one, and the report comes back with a very high similarity score against a single source your instructor cannot open, because papers held at other institutions are shown as a match without the text. What they can see is that a large block of your submission already exists in the repository, attributed to a submission at a named institution. If the earlier paper was yours, the explanation is easy and completely credible. It is much easier to give before someone asks for it.
Note also what this does to your number. A high score here is not evidence of anything on its own, which is the general point of what a good Turnitin similarity score is: the score counts matched text and knows nothing about who wrote it or whether it is cited.
What you can reuse without asking anyone
The rule covers submitted work. It does not cover the research that produced it, and the distinction is generous once you see it.
- Your sources. A bibliography is not the work. If you spent three weeks finding the right twenty papers on urban heat and cooling demand, those twenty papers are yours to use again, and nobody expects you to find different ones.
- Your notes and your reading. Anything you never submitted for credit is just knowledge you have.
- Your topic. Writing about the same subject twice is normal, and in a degree with any coherence it is unavoidable. What has to be new is the work.
- Your own published work. If your earlier piece was published rather than submitted for credit, it is a citable source like any other. Oxford's policy says exactly this: "Where earlier work by you is citable, i.e. it has already been published, you must reference it clearly."
- Your data, usually. Reusing a dataset you collected is common and generally fine with disclosure, but this one genuinely varies by discipline, so ask.
So the honest version of "I already did this" is usually to keep the reading and write the paper again. That is a smaller job than it sounds, because the expensive part of a paper is the search and the reading, not the drafting. If you are rebuilding an argument you have made before, finding a source for a claim you already wrote covers the reverse direction, which is the position you are in when the argument exists first.
How to ask, and how to cite yourself
Asking is short, and it works better when it is specific about what is carried over. Two sentences:
I wrote a paper on congestion charging for [module] last year, and this assignment overlaps with about a third of it, mainly the literature on price elasticity. Would you be willing to let me build on that section, and if so, how would you like me to mark what is reused?
That does three things a vaguer request does not. It names the overlap rather than asking a general permission question, it makes clear you are extending rather than resubmitting, and it hands the instructor the decision about how to signal it. Ask by email, so the answer is in writing.
If permission is granted, cite the earlier work. In APA, an unsubmitted or unpublished piece is treated as an unpublished manuscript, with the status in square brackets after the title, following the pattern APA sets out for unpublished works:
Ortiz, R. (2025). Congestion charging and travel behaviour in mid-sized cities [Unpublished manuscript]. Department of Geography, University of Leeds.
In-text it behaves like any other author-date citation. MLA and Chicago both have equivalent forms for unpublished work; the principle is the same in all three, which is that your earlier paper appears as a source rather than as invisible text. Citing yourself feels strange the first time. It is the thing that converts a rule violation into an ordinary reference, which is a good trade for one line in a bibliography.
If you already handed it in
If you submitted reused work without permission and have realised what that means, the useful order is: read your institution's policy so you know which rule you are talking about, then tell the instructor before the similarity report does. What you say is short and does not require self-flagellation. You reused work from a previous assignment, you did not realise permission was needed in advance, here is what came from where.
Outcomes for a first, self-reported instance of this are usually at the mild end, and the codes above are written to be graduated rather than punitive. The version that goes badly is the one discovered by a report and then denied, because at that point the conversation is no longer about a reused essay. Our wider walkthrough of that conversation is in how to avoid plagiarism, and if the essay in question also involved an AI tool, be straight about that at the same time rather than in a second conversation a week later. What to actually write in an AI disclosure has the wording.
The general habit worth building, whichever of these situations you are in: keep the history of what you wrote and when. A draft folder with dates, or an editor that keeps its own version history, turns "I wrote this myself in 2024" from a claim into something you can show. CiteOwl keeps a full revision history of every document and a record of which sources each claim came from, which is useful for exactly this and, more often, for the ordinary case where you need to remember why a paragraph says what it says.