BMJ Open Formatting Requirements: Complete Author Guide

Here is the part most guides get wrong. For a research article, BMJ Open's current author page gives no word limit, and it sets no word cap on the abstract either. The journal asks for a manuscript that is "as succinct as possible" and puts its real limits elsewhere. The 4,000-word figure quoted widely (older versions of this page included) is not on the author page as of September 2026. Treat it as a sensible ceiling, not a rule, and if the ScholarOne form asks for a word count, give the real one.

Source: BMJ Open instructions for authors, submission guidelines by article type (accessed October 8, 2026)

The eight-item display cap is the one that bites. BMJ Open says going past it hurts readability and may ask you to move material into supplementary files, so plan which tables live in the main text before you write the Results.

Protocol papers are a distinctive feature of BMJ Open. Publishing your study protocol before the data are in provides a citable record of your planned methods and helps prevent outcome reporting bias. Timing matters: BMJ Open will not consider a protocol once data collection is complete, and protocols near the end of collection are judged case by case.

BMJ Open requires a structured abstract for research articles. The author page lists the headings but no word cap.

For Original Research:

• Objectives

• Design

• Setting

• Participants

• Interventions (delete if there were none)

• Primary and secondary outcome measures (quantitative studies only)

• Results

• Conclusions

• Trial registration (registry and number, for clinical trials)

This is more granular than most journals. Where other journals combine design information into a single "Methods" heading, BMJ Open separates Design, Setting, Participants, and Interventions into individual sections.

Objectives should state exactly what the study aimed to do. One sentence is usually enough.

Design specifies the study type: randomized controlled trial, prospective cohort, cross-sectional survey, etc. Include the time period.

Setting describes where the study was conducted: primary care, hospital, community, country.

Participants includes the number enrolled, key demographics, and inclusion criteria.

Results should report the main outcomes with 95% confidence intervals, the exact P value where it matters, and numbers needed to treat or harm when they apply. BMJ Open also asks for absolute rather than relative risks whenever possible, so an abstract that gives only a hazard ratio is under-reported by the journal's own standard.

Conclusions should be limited to what the data support. BMJ Open reviewers are particularly attentive to conclusions that overreach the data, and the open review process means any reviewer criticism of your conclusions will be publicly visible.

BMJ Open requires a "Strengths and limitations of this study" section immediately after the abstract, published as a summary box. The rule is tighter than most authors assume: no more than five bullet points, and they must relate specifically to the methods, not the results.

This is unusual. Most journals ask authors to discuss limitations in the Discussion section, but BMJ Open puts them front and center. Don't treat this as a throwaway. Reviewers compare your self-assessment against their own evaluation, and a paper that omits obvious limitations will draw criticism in the open review.

Good Strengths and Limitations statements are specific. "Large sample size" is vague. "Population-based cohort of 45,000 participants with 10 years of follow-up" is informative. Similarly, "potential for confounding" is generic. "Unable to adjust for socioeconomic status due to data limitations" is honest and specific. A bullet like "Treatment reduced admissions by 20%" breaks the rule outright, because it reports a finding; move it to the abstract.

Two BMJ rules change how you build the title page. First, the published byline comes from the author fields you fill in on ScholarOne, not from the Word file, so a typo in the submission form is the one that gets printed. Second, BMJ Open mandates an ORCID iD for every submitting author.

What the manuscript itself should carry:

• A title that names the research question and the study design, without announcing the result

• Author names and institutional affiliations, matching the ScholarOne entries

• Corresponding author's contact details

• Keywords chosen as MeSH descriptors (BMJ points authors to MeSH on Demand and the MeSH Browser)

BMJ Open also requires:

• Contributor statement: each author's contribution listed individually at the end, following ICMJE authorship criteria

• Funding statement: BMJ Open suggests the exact wording, either "This work was supported by [funder] grant number [xxx]" or "This research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors"

• Competing interests statement: covering all authors

• Patient and public involvement statement: placed in the Methods section under the subheading "Patient and public involvement", saying how patients or the public shaped the question, design, conduct, or dissemination

The PPI statement applies to research articles, study protocols, and cohort profiles. If patients weren't involved, say so directly; BMJ Open still considers those papers, and it will ask for the statement if it is missing. Don't fabricate involvement.

BMJ Open caps research articles at eight figures and tables combined. Count before you upload: a flow diagram and a baseline characteristics table already use two of the eight.

Figure requirements:

• Minimum resolution for non-vector files (TIFF, JPEG): 300 dpi, rising to 1,200 dpi for line art

• Accepted formats: TIFF, EPS, JPEG, or PDF; TIFFs must be single-layer flat files, and text in EPS files should be outlined

• A multipanel figure goes up as one composite file; BMJ will not accept a figure split across several files

• Colour is free: BMJ Open has no colour charges

• Each figure uploaded as a separate file, never embedded in the manuscript, with legends at the end of the Word file

• Figures run through automated quality control, and anything below the minimum standard comes back for resupply

Table requirements:

• Built in Word and placed in the main text where the table is first cited, not collected at the end

• Cited in numerical order, with a header at the top of each table

• No tables embedded as Excel objects; paste Excel tables into the Word file as editable tables

• Data in a table should not be repeated in the text or figures

• Any table longer than 2 pages is published as online-only supplementary material

For systematic reviews, BMJ's reporting-guidelines page lists the PRISMA checklist, flowchart and structured abstract checklist, and BMJ Open asks for completed checklists, structured abstracts and flow diagrams to be uploaded with the submission.

BMJ Open uses BMJ's Vancouver reference style. References are numbered consecutively as they appear in the text and matched to a numbered list at the end.

Key formatting rules:

• In-text citations go in square brackets [6] or as superscripts; parentheses are for in-line lists

• Separate multiple citations with commas [6, 9, 12] and use a hyphen for a run [12-15]

• List the first three authors by last name and initials, then "et al" if there are more

• Journal titles abbreviated per NLM/Medline standards

• Include volume, page range (or article number for e-publications), and year

• Each entry needs at minimum the author names, the article title, and a DOI or PubMed ID

• References cited only in a table or figure legend still go in the main reference list

Example reference:

1 Smith AB, Jones CD, Williams EF, et al. Patient engagement in primary care quality improvement: a cluster randomized trial. BMJ Open 2025;15:e098765. doi:10.1136/bmjopen-2025-098765

Note that BMJ Open uses "e" article numbers rather than page ranges for its own articles. When citing other BMJ Open papers, use the article number format.

Zotero and EndNote both offer BMJ output styles, and the BMJ Author Hub reference section shows the target format. Check the three-author rule in your export: generic Vancouver styles often list six.

Worked reference-manager check:

Wrong author-date export: Smith, A. B., Jones, C. D., Williams, E. F. Patient engagement in primary care quality improvement. BMJ Open. 2025;15:e098765.
BMJ Open target: Smith AB, Jones CD, Williams EF, et al. Patient engagement in primary care quality improvement: a cluster randomized trial. *BMJ Open* 2025;15:e098765. doi:10.1136/bmjopen-2025-098765

The common failure is not the article number itself. It is a reference-manager export that leaves APA-style author punctuation, Word footnotes, or EndNote field codes in the manuscript. BMJ Author Hub says references should be part of the main text document, not inserted as Word footnotes or live citation fields.

BMJ Open publishes supplementary material online alongside the article. BMJ publishes it as supplied, without copyediting, typesetting, or proofing, so an error in a supplementary table goes live exactly as you uploaded it. Upload each file under the "Supplementary File" designation and cite it in the main text.

Supplementary material can include:

• Additional tables and figures

• Extended methods and statistical analyses

• Questionnaires and survey instruments

• Interview topic guides (for qualitative research)

• Data collection forms

• Additional sensitivity analyses

BMJ's Author Hub says supplemental material may be submitted in PDF format, files should not exceed 350MB, and material hosted on a third-party website or platform is not accepted.

For systematic reviews, the full search strategy for each database can go in supplementary material, and for qualitative studies so can the interview guide or focus group schedule.

This one is not a preference. BMJ's formatting rules say the manuscript must be submitted as a Word document, and a PDF is not accepted. The journal runs on ScholarOne, and its submission prefill tool reads the title, abstract, authors, institutions, and funders straight out of the Word file.

Word file rules BMJ states:

• Order: abstract, main text, tables in place, then acknowledgements, competing interests, funding, and other statements, then references

• Heading hierarchy: BOLD CAPS, bold lower case, plain text, italics

• No references in the abstract

• Remove hidden text, headers, and footers before upload

• SI units throughout, except blood pressure in mm Hg

If you write in LaTeX:

• Convert to Word before submission; a compiled PDF will not be accepted as the manuscript file

• Rebuild tables as native Word tables after conversion, since converted tables often arrive as images or broken cells

• Re-check the reference list for leftover BibTeX keys or field codes

For most BMJ Open submissions (clinical trials, observational studies, public health research), Word is the only route anyway.

Use BMJ's current formatting your paper page as the starter-file reference before upload. The practical template is not a decorative .docx; it is a Word manuscript that follows BMJ's order, heading hierarchy, table handling, reference handling, and separate upload-file requirements.

BMJ Open's open peer review policy means:

• Reviewer names are published alongside accepted articles

• The review history (previous manuscript versions, reviewer reports and author responses) is published online

• Reviewers know their identity will be revealed if the paper is accepted

• Authors should write response letters knowing they will be public

This has practical implications for formatting. Your response to reviewers should be professional, detailed, and well-organized. Use a point-by-point format addressing each comment. These responses become part of the permanent published record.

BMJ Open has several unique requirements beyond the standard formatting rules.

1. EQUATOR checklists are non-negotiable. Upload the checklist for your design, or say why none applies. A missing checklist is a return-to-author problem that is entirely within your control.

2. Strengths and Limitations box is mandatory. This appears immediately after the abstract and is separate from the Discussion section. Up to five bullets, all about methods.

3. Patient and Public Involvement statement is required. Even if patients weren't involved, you must include a statement. This is increasingly common across BMJ journals.

4. Open review changes the dynamic. Knowing that reviews will be published tends to produce more constructive and thorough reviews. But it also means your response letters need to be polished. Typos and terse responses will be permanently visible.

5. Protocol papers are a strong option. If you're planning a clinical trial or systematic review, consider submitting the protocol to BMJ Open before starting data collection. This establishes your methods publicly and can prevent accusations of outcome switching.

The most frequent triggers for administrative return at BMJ Open:

• Missing EQUATOR checklist, with no explanation of why none applies

• Missing Strengths and Limitations box

• Strengths and limitations bullets that report results instead of methods

• Missing Patient and Public Involvement statement

• Abstract using wrong headings (e.g., "Methods" instead of "Design")

• Checklist page numbers that don't match the manuscript

• More than eight figures and tables in the main manuscript

• Manuscript uploaded as a PDF instead of a Word file

EndNote, Zotero, Paperpile, and Word can all make the file look polished while leaving BMJ Open-specific issues unresolved. The highest-risk pitfall is a reference list that displays numbered citations but still contains Word citation fields, EndNote codes, APA-style punctuation, or footnote/endnote references. BMJ's formatting guidance expects references inside the main text document, not as footnotes or hidden citation objects.

The second pitfall is treating the EQUATOR checklist as an attachment you fill in after writing. BMJ Open reviewers and editors can compare the checklist page numbers against the abstract, methods, figures, tables, and supplementary files. If the checklist says an item is on page 7 but the method, sample-size, ethics, or outcome-definition text is missing, the formatter did not solve the submission problem.

The third pitfall is table cleanup. Word tables need editable cells, cited table numbers, column headers, and readable footnotes. A pasted image of a table, a shaded spreadsheet screenshot, or a vertical-line-heavy table can survive a local formatting pass and still create production or review friction.

In our pre-submission review work on BMJ Open submissions, the formatting failures that matter are rarely cosmetic. They usually sit in the manuscript package: abstract headings, reporting checklist upload, ethics statement, PPI statement, data-availability statement, table files, and reference-manager cleanup. Manusights submission analysis treats these as specific failure pattern signals because editors routinely flag the missing file or statement before they evaluate the clinical argument. In practice, we see authors fix fonts and margins while leaving the checklist, ethics, and data-availability evidence misaligned. That is why we check the Word file, supplementary files, and submission answers together instead of treating formatting as a final visual pass.

Reporting checklist not submitted or in the wrong format. BMJ Open asks for completed reporting checklists for the study types where a guideline applies: CONSORT for randomized trials, STROBE for observational studies, PRISMA for systematic reviews and meta-analyses, and STARD for diagnostic accuracy studies. The checklist must be submitted as a separate file at the time of submission. Manuscripts that describe a clinical or observational study without attaching the appropriate checklist are returned for correction.

Strengths and limitations bullets that report findings. BMJ Open limits the box to five bullets about how the study was done. Drafts moved over from another journal's template tend to carry a "key findings" bullet into it, or open with the headline effect size. Those bullets have to come out before upload, and the methods point they displaced (the unmeasured confounder, the single-site sample) is usually the one BMJ Open reviewers were going to raise in public anyway.

Ethical approval statement absent or insufficient. BMJ Open requires that all studies involving human participants include a statement of ethical approval specifying the name of the ethics committee, the number or ID of the approval, and that participants gave informed consent. Manuscripts that state "the study was approved by an ethics committee" without specifying the institutional review board name and reference number are returned for clarification. For studies using data from registers or records that are exempt from formal review, a statement explaining the exemption is required.

Patient and Public Involvement (PPI) statement missing. BMJ Open requires research articles, study protocols and cohort profiles to include a Patient and Public Involvement statement that explains how patients or members of the public were involved in the design, conduct, reporting, or dissemination plans for the research. If patients were not involved, the statement just needs to say so. Manuscripts that do not address PPI in any form are asked to add it, and it belongs in the Methods under its own "Patient and public involvement" subheading.

Open data statement not included or data not deposited. BMJ Open strongly encourages open data sharing and requires at minimum a data sharing statement that explains what data are available, where they are available, and any conditions under which they can be accessed. If the data cannot be shared (for ethical or legal reasons), this must be explicitly stated with a justification. Manuscripts without any data availability statement are asked to add one before review.

A BMJ Open submission readiness check evaluates manuscript structure, reporting checklist compliance, ethical approval documentation, and PPI statement against these desk-rejection patterns before you submit.

For a manuscript-specific signal before you submit, run a BMJ Open submission readiness check.

Or see example reports before you finalize.