Acceptable Similarity Percentage for Journal Submission

Similarity thresholds are one of the most misunderstood aspects of journal submission. Many researchers either don't know their target journal's threshold or rely on outdated information. This guide summarizes commonly stated similarity expectations based on publicly available publisher author guidelines. Thresholds vary by journal and change over time, so always verify against your specific target journal.

Similarity Thresholds by Publisher

What "Similarity" Actually Means

A similarity score represents the percentage of your manuscript text that matches content in iThenticate's database. Critically, similarity is not the same as plagiarism — it is a computational measurement. A high similarity score can result from properly cited quotations, shared technical terminology, or self-citation of your own prior work.

Publishers interpret similarity scores contextually. An editor reviewing a 20% similarity score will look at the match breakdown: is it one large block matching a single source (suspicious), or is it distributed across many small matches to properly cited papers (acceptable)?

What Is Excluded From the Similarity Score

Most publishers configure iThenticate to exclude the following by default:

• Reference list and bibliography
• Properly formatted quotations (with quotation marks)
• Matches of fewer than 8 consecutive words
• Author name and affiliation blocks

Similarity Targets at Some High-Impact Journals

The targets below reflect the common <15% working figure cited across major publishers. Specific journals may state different requirements in their author guidelines, so confirm before submitting.

How to Reduce Your Similarity Score

If your similarity score is above your target threshold, these sections typically contain the most fixable matches:

1. Introduction and literature review — rewrite descriptions of other researchers' work using your own framing and vocabulary
2. Methodology — if using a standard instrument or validated scale, paraphrase the description rather than copying from the original paper
3. Self-plagiarism — if you're reusing your own prior methodology section, rewrite it substantially or add a proper self-citation disclosure
4. Boilerplate academic phrases — "this study investigates", "the results indicate that", "as shown in Table X" — vary these standard constructions

Single-Source Matches: The Biggest Red Flag

A total similarity of 18% distributed across 30 sources is far less concerning to an editor than 18% similarity where 12% comes from a single source. Single-source concentration — where a large block of your text matches one specific paper — is the pattern most likely to trigger a plagiarism investigation, regardless of total percentage.

Most iThenticate reports show you the match breakdown by source. Always review this breakdown, not just the total percentage.

Conclusion

Target below 15% total similarity for most Q1 journal submissions, below 10% for high-impact journals, and always review the per-source breakdown in addition to the total score. Use ScholarAI to identify high-risk passages before your final iThenticate check.