JCI Formatting Requirements: Complete Author Guide

JCI uses a single-paragraph abstract without subheads for Research manuscripts. Clinical Research and Public Health manuscripts use a structured abstract with Background, Methods, Results, Conclusion, Trial registration, and Funding sections.

• Word limit: 200 words maximum for Research; 250 words maximum for Clinical Research and Public Health

• Structure: Single paragraph, no subheads (Research); structured sections (Clinical Research and Public Health)

• Citations: The abstract carries no primary data or references

• Keywords: The submission system asks for keywords; JCI lists them among the items to have ready

The 200-word cap is tight. You need to state the problem, describe the approach briefly, and present the main findings. Don't spend 40 words on background context that any reader of JCI already knows. Start with the specific gap or question.

One formatting detail that trips people up: JCI asks you to define all nonstandard abbreviations at first use in the abstract. Standard JCI abbreviations and acronyms (JCI publishes a list) may be used without definition; all others are defined at first mention in the text, with the abbreviated form in parentheses.

JCI limits the main text of a Research manuscript to 9 display items in total, with figures and tables combined. A figure may have no more than 10 parts (A to J).

Figure specifications:

Table requirements:

• For revised manuscripts, tables must be in Word table format, not pasted in as an object from another application

• Every column needs a header

• Each table needs a brief descriptive title and must fit on a single page (no more than 50 rows in the main text)

• Footnotes use uppercase superscript letters (A, B, C), ordered row by row

Supplemental figures: For Research and Clinical Research and Public Health manuscripts, JCI places no limit on the number of display items in the supplemental material, but editors and reviewers will push back on excessive supplemental data that should either be in the main paper or cut entirely. Research Letters are limited to 4 supplemental display items.

Multi-panel figures are the norm at JCI. A single figure with panels A through H is common, and it counts as one display item toward your limit of 9. Use this to your advantage. Group related experiments into composite figures rather than spreading them across separate numbered figures.

One JCI-specific detail: a graphical abstract is recommended but optional. It is a single image that visually represents the article's primary findings, and published graphical abstracts appear at the top of the page in the online version of the article.

JCI uses a numbered sequential citation system. References are cited as numerals in parentheses in the text and listed numerically in the reference list.

In-text citations: Numerals in parentheses preceded by a space (e.g., "as described previously (1, 2)"). Numbers are assigned in the order references first appear in the text. Do not use superscript, brackets, italics, or other formatting for citations.

Reference list format:

1. Smith AB, et al. Title of the article in sentence case. J Clin Invest. 2025;135(4):e12345.

Key formatting details:

• Author names: Last name followed by initials with no periods (e.g., "Smith AB")

• With 3 or more authors, list only the first author followed by ", et al."

• Journal names follow standard NLM abbreviations

• Include issue numbers in parentheses

• Page numbers or article IDs are required

• Do not include extraneous information such as PubMed identification (PMID)

JCI doesn't publish a reference cap for Research manuscripts, but references count toward the 12,000-word maximum. If you're at 80+ references for a Research manuscript, that's a signal your paper might be trying to do too much.

JCI's own abbreviation: When citing JCI itself, use "J Clin Invest" as the abbreviated journal name.

Unlike the publisher-wide reference templates that Cell Press, Elsevier, or Wiley apply across a whole portfolio of journals, JCI's reference requirements differ in a way that trips up reference managers: the journal instructs that citations in the text must not use superscript, brackets, italics, or other formatting. If you import a default output style, the citation numbers will often arrive as superscripts, and they must be changed to numerals in parentheses before submission. JCI staff strongly encourage EndNote in Microsoft Word, and staff may return a submission if references are not prepared in Journal format.

JCI calls supplementary content "Supplemental Material" (not "Supplementary" with a y, which is a small but notable style choice). Supplemental material is published alongside the accepted article.

Supplemental material should be organized as:

1. Supplemental Methods: Extended methodological details that don't fit in the main Methods section

1. Supplemental Figures: Numbered sequentially (Supplemental Figure 1, Supplemental Figure 2, etc.)

1. Supplemental Tables: Numbered sequentially; spreadsheets of large tables are uploaded separately

1. Supporting Data Values: A single XLS file, separate from the supplemental PDF, with the values behind every graph and reported mean, one tab per figure panel

Supplemental material (except videos and spreadsheets) should be combined into a single PDF. Videos (in MP4 format) and large spreadsheets are uploaded as standalone files, and JCI does not count large data sets as supplemental material.

There's no limit on the number of supplemental display items for Research and Clinical Research and Public Health manuscripts, but the supplement shouldn't become a second paper. If your supplemental data is longer than your main manuscript, reconsider whether some of that data belongs in the main figures or whether it's truly necessary.

Data deposition: JCI requires that large gene expression microarray, high-throughput sequencing, and proteomics data sets be deposited in appropriate public repositories (GEO, PRIDE, etc.) with accession numbers in the Data availability section; deposition of other large data sets is strongly encouraged.

JCI requires both a title page (as the first page of your manuscript) and a cover letter (submitted separately).

Title page must include:

• Title of no more than 15 words, including conjunctions; no subtitles, colons, periods, or nonstandard abbreviations

• Full names of all authors, with no titles, honorifics, degrees, or certifications, and their affiliations

• Corresponding author contact details

• Conflict of interest statement

Cover letter should address:

• Why the work fits JCI's scope (translational biomedical research)

• What's new compared to existing literature

• Any potential conflicts of interest

• The 3 potential reviewers JCI asks for, and up to 2 reviewers to exclude (optional), with reasons

JCI's editorial scope is translational medicine. Basic science papers without a clear path to clinical relevance don't fit, and the cover letter is where you establish that connection. Don't just describe what you found. Explain what it means for understanding or treating disease.

These are the details that experienced JCI authors know but that first-time submitters often miss:

Methods are in the main text. Unlike Nature, Cell, and Science, JCI keeps the Methods section within the body of the paper, after the Discussion. The order is Introduction, Results, Discussion, Methods, and JCI's instructions present the main text in that order.

No structured significance statement. Unlike PNAS, JCI doesn't require a separate lay-audience summary or significance statement. The abstract serves this purpose.

Author contributions statement. Required for Research and Clinical Research and Public Health manuscripts. Must specify the contribution of each author, identified by initials. It appears in the Manuscript information section, before the references.

Conflict of interest statement. Mandatory, even if there are no conflicts to declare (JCI gives a standard sentence for that case). Any potential conflict of interest by any author must be stated on the title page.

Ethics statements. Human and animal studies need a stand-alone Study approval paragraph near the end of Methods, naming the official institutional review board(s) and their location; for human studies, include a statement that written informed consent was received.

ORCID iDs. JCI recommends providing ORCID information for authors in the submission system.

Abbreviation use. Standard JCI abbreviations and acronyms, such as DNA, RNA, PCR, and ELISA, may be used without definition; all others are defined at first mention, with the abbreviated form in parentheses.

Title formatting. JCI titles must be clear, concise, and limited to 15 words including conjunctions, and they may not use subtitles, colons, periods, or nonstandard abbreviations. They should refer to the relevant disease or disease model studied.

JCI staff may return a submission that does not follow the Journal's formatting requirements, and for Research Letters and Letters to the Editor JCI states that revisions will be returned if word limits and figure size are not met. Reference managers like EndNote and Zotero often output the wrong citation style for JCI (superscript numbers rather than numerals in parentheses). Work through the details below manually rather than trusting a one-click style.

1. Each section starts on a new page. JCI's instructions ask that each section of the manuscript begin on a new page.

1. Double spacing. The manuscript must be double-spaced throughout, including references and tables. Figure legends may be single-spaced if necessary to keep a figure and its legend on the same page.

1. Page numbers. All pages are numbered.

1. Figure legends. In the first-submission PDF, each figure should be shown with its legend on the same page. In the revised-submission Word or RTF file, figure legends are keyed in as text after the references, without the figures, and each legend is limited to 300 words.

1. Statistical reporting. All statistical tests must be identified in the Methods and figure legends. Define the P value used to determine significance, report the exact number of samples (n) for each figure panel representing multiple experiments, and define error bars. Do not show variance or statistics for fewer than 3 independent samples.

JCI's formatting requirements are more forgiving than many top-tier journals. The 9,000-word recommendation (12,000 maximum) gives you room to tell a complete story, the 9-display-item cap is workable, and the Introduction, Results, Discussion, Methods structure is standard enough that most authors don't need to restructure their manuscripts.

The bigger challenge at JCI is meeting the editorial bar for translational impact. If you want to check whether your manuscript is ready for a high-impact submission, JCI pre-submission checklist to catch the technical issues that lead to desk rejection.

For the most current author guidelines, check JCI's instructions for authors. Checklist documents for revised submissions are available through that page.

If you're also evaluating journal fit, our guides on JCI impact factor and PNAS formatting requirements cover journals with a similar translational scope.

For manuscripts targeting Journal of Clinical Investigation, four patterns generate the most consistent desk-rejection outcomes.

Novelty does not meet JCI's bar for "a major conceptual advance." JCI publishes mechanistic studies with direct clinical implications, and its editors apply a very high novelty bar: the work must provide genuinely new insight into the mechanism of a disease process or therapeutic response, not simply add an observation to existing knowledge.

Papers that characterize a new molecular marker without mechanistic resolution, or that extend a known pathway to a new cell type without clinical implication, are desk-rejected. The cover letter must state the transformative insight explicitly in one to two sentences.

Translational bridge absent between mechanism and human relevance. JCI is a translational journal. Mechanistic papers that work entirely in cell lines or mouse models without patient sample validation, genetic association in a human cohort, or a direct link to a clinical trial or therapeutic target are desk-rejected. JCI's stated criteria for external review ask for a relevant in vivo model and/or a primary human cell or tissue model, and human tissue data or patient correlation makes the translational case much stronger.

Data availability and source data not provided. JCI requires a Supporting Data Values file (a single XLS file, separate from supplemental material, with the values underlying each graph and reported mean, one tab per figure panel) for new submissions with data shown in graphs or as mean ± standard deviation. The Data availability paragraph must also say how supporting analytic code can be accessed (for code deposited in GitHub, the commit URL and ID).

Conflict of interest disclosures incomplete for clinical authors. JCI's high-impact clinical readership means editors scrutinize COI declarations carefully. Authors with relevant industry relationships (consulting, equity, advisory roles, honoraria) must disclose these specifically. Generic "no conflicts" declarations where relevant relationships exist, or incomplete declarations that emerge during review, damage the publication relationship and can cause post-acceptance delays.

A JCI submission readiness check evaluates manuscript scope, translational bridge quality, and source data compliance against these desk-rejection patterns.