Direct answer
Turnitin similarity detection compares your submitted document against a large repository of student papers, web pages, and academic publications, then reports the share of your text that matches existing sources as a similarity score [1]. The score is a matching measurement, not a plagiarism verdict: it tells you where text overlaps, and a human still has to decide whether that overlap is properly cited or not [1]. Understanding what is compared, what the percentage counts, and what your institution considers acceptable is what turns a raw number into a usable pre-submission check.
Introduction
Turnitin similarity detection compares your submitted document against a large repository of student papers, web pages, and academic publications, then reports the share of your text that matches existing sources as a similarity score [1]. The score is a matching measurement, not a plagiarism verdict: it tells you where text overlaps, and a human still has to decide whether that overlap is properly cited or not [1]. Understanding what is compared, what the percentage counts, and what your institution considers acceptable is what turns a raw number into a usable pre-submission check.
How does Turnitin similarity detection actually work?
Turnitin processes a submission in stages. The document is parsed and broken into text segments, and each segment is compared against the repository contents the system holds at that moment [2]. Those contents include current and archived student papers previously submitted to the institutional database, a large index of internet pages, and periodical and publication content [2]. Matches are then aggregated into a Similarity Report where every detected overlap is highlighted and linked to its source, so you can see exactly which passage matched what [1][2].
Two details shape how you should read the output. First, the report is generated after the submission has been processed, so a draft uploaded minutes before a deadline may not return a finished report immediately [2]. Second, the report is a match map rather than a judgement — Turnitin explicitly frames the Similarity Report as a tool for interpreting overlap, and the interpretation step belongs to the instructor or the author [1]. A single 8% match to one paper and eight separate 1% matches to eight different papers produce the same headline number but very different conversations.
The repository question also matters for your own workflow. When a paper is submitted to the standard institutional database, it can become part of the pool that future submissions are compared against, which is why students are often warned that a draft check can create a match for a classmate [2]. That is a repository effect, not a detection effect, and it is the reason non-repository checking exists as a separate option for pre-submission drafts.
What similarity score is safe, and what does the percentage really mean?
There is no universal safe number. Turnitin does not publish a pass mark, and institutions and individual instructors set their own acceptable ranges for assignments, which is why the same essay can be unremarkable in one department and flagged in another [3]. What the percentage represents is the proportion of your document's text that matched a source in the repository — it is a coverage figure, not a severity figure [3]. A 30% score built almost entirely from correctly quoted and referenced material is a different document from a 12% score built from unattributed paraphrase.
In practice, reviewers tend to look at composition rather than the headline figure. Reports in the low-to-mid range are common for well-referenced academic writing, while higher scores usually prompt a closer read of the individual matches and their context [3]. Many institutions also configure the report to exclude bibliography and quote sections, or to ignore matches below a set word count, which can shift the displayed percentage without changing a single word of your text [3]. If your score looks surprising, check those settings before you panic.
The most useful habit is to read the match list, not the number. Open each highlighted source, confirm whether the overlap is a direct quotation with a citation, a correctly attributed paraphrase, or an unattributed lift, and fix only the third category [1][3]. That process also protects you from the opposite error — rewriting original analysis simply because a common phrase triggered a small match, which weakens your writing without improving integrity.
Can you check your own similarity score before you submit?
Only if someone enables it for you. Turnitin does not offer a general public self-check portal, and a student can typically see a Similarity Report only when the instructor has turned on student access for that specific assignment [4]. Where that setting is enabled, you may be able to submit a draft and view the report before the final deadline, but availability, attempt limits, and whether the draft enters the paper repository are all controlled by the instructor's assignment configuration [4]. That is why the ability to self-check varies so widely between modules at the same university.
This creates a real planning problem. If your module does not offer a student preview, you reach the deadline with no visibility into which passages will match, and the first time you learn about a 28% score may be after the submission is locked [4]. Students in that position usually want three things: the actual similarity percentage, the highlighted matches with sources, and the AI writing signal that now accompanies most reports — all before the final upload rather than after it.
The practical workaround is a pre-submission check that mirrors the institutional report without adding your file to the student paper database. That is the specific gap turnitin0 was built to close.
If your module does not give you a student preview, you do not have to submit blind. turnitin0 returns the same two PDFs your instructor sees — a similarity report with highlighted matches and a source list, plus the AI writing report — from a non-repository check, so your draft is not added to the student paper database and cannot create a match for anyone else. Most orders finish in 5–15 minutes, which is fast enough to act on the findings while you can still edit.
※ Turnitin0.com - Actual Turnitin AI Report Cover, Score, Flag And Similarity Summary
FAQ
Does a high similarity score mean I plagiarised?
No. The score measures matched text coverage, and matches can come from correctly cited quotations, standard terminology, or a shared reference list [1][3]. The decision about intent belongs to your instructor, which is why the match list matters more than the number.
Why does my score change between two checks of the same essay?
The repository grows continuously, so a later check compares your text against more student papers and web pages than an earlier one did [2]. Report settings such as excluding quotes or bibliography can also change the displayed percentage [3].
Can I see my similarity report before the deadline?
Only if your instructor enables student access for that assignment; Turnitin has no general public self-check portal [4]. Where the setting is off, an independent non-repository check is the usual alternative.
Does checking a draft add my paper to the student database?
With a standard institutional submission it can, depending on the assignment settings, which is why draft checks carry a risk for classmates [2][4]. A non-repository check is designed to avoid adding the file to that pool.
What should I do if one passage is flagged and I did cite it?
Open the match, confirm the citation is present and correctly formatted, and leave the wording alone if it is a legitimate quotation [1]. Fixing only genuinely unattributed overlap keeps your analysis intact.