
How to Write a Research Manuscript: Complete IMRaD Format Guide
A practical walkthrough of the IMRaD format for research manuscripts, covering the smartest writing sequence, key section strategies, and how to sidestep the pitfalls that lead to rejection.
Crafting a research manuscript is one of the most demanding milestones in any academic career. Structure and clarity can be the deciding factor between acceptance and a desk rejection, and with journals turning away as many as 70% of all submissions, building solid manuscript skills is not optional.
This guide covers the IMRaD framework from the ground up, explains the smartest order in which to draft each section, and highlights the patterns that most often doom a paper before it reaches peer review.
Understanding the IMRaD Framework
IMRaD is an acronym for Introduction, Methods, Results, and Discussion. The format became the dominant structure for scientific manuscripts during the 1970s and is now endorsed by bodies such as the American Psychological Association and the International Committee of Medical Journal Editors.
Its appeal lies in how naturally it mirrors the scientific method itself:
| Section | Core Purpose | Central Question |
|---|---|---|
| Introduction | Background and rationale | Why was this study necessary? |
| Methods | Procedures and design | How was the research conducted? |
| Results | Findings and data | What did the study uncover? |
| Discussion | Interpretation and context | What do the findings actually mean? |
Why IMRaD endures: Readers, reviewers, and editors can navigate a predictable structure efficiently, locate the information they need, and assess the quality of your work more fairly.
The Counterintuitive Drafting Sequence
A key insight that separates productive manuscript writers from frustrated ones: the best order to draft a manuscript is not the order in which it gets read.
Experienced scientific writers consistently recommend a sequence that starts at the data and builds outward:
Recommended Drafting Sequence
- Tables and Figures - Lay out your data story before anything else
- Methods - Document what you did while it is still fresh
- Results - Narrate what the figures and tables show
- Discussion - Interpret and contextualize your findings
- Introduction - Frame what you already know the paper delivers
- Conclusion - Distill the essential takeaways
- Abstract - Summarize the complete, finished manuscript
- Title - Refine it once the full scope is clear
The Logic Behind This Approach
Opening with figures and tables forces you to confront the actual story your data tells. As Redwood Ink notes, "The first task is to lay out your tables and figures in the best order. This order will likely be different from the order in which you carried out the experiments." Knowing your visual narrative first keeps every other section anchored to reality.
Tackling Methods early is simply practical. You know precisely what was done, so the prose flows with little resistance. It also surfaces any documentation gaps before you have invested hours in other sections.
Postponing the Introduction until near the end is the most counterintuitive move, but it prevents a pervasive problem: introductions that promise more than the paper ultimately delivers. Writing it last ensures your framing matches your actual contribution.
Drafting Each IMRaD Section
Title: The Entry Point
A strong manuscript title is:
- Specific: Signal the key finding or focus area
- Concise: Aim for roughly 10 to 15 words
- Searchable: Incorporate terms your target audience actually uses
- Accurate: Represent neither more nor less than the paper contains
Illustrative contrast:
- No: "A Study on Cancer Cells" (too broad to be meaningful)
- Yes: "CRISPR-Cas9 Editing Reduces Tumor Growth in BRCA1-Mutant Breast Cancer Models"
Abstract: Your 250-Word Pitch
Reviewers and busy readers often go no further than the abstract. According to Frontiers author guidance, a solid abstract should:
- Cover the complete study in 250 to 300 words
- Read as a self-contained summary requiring no reference to the full paper
- Touch on: background, objective, methods, results, and conclusion
- Omit citations and unexplained abbreviations
Draft the abstract last. Only then can it accurately reflect what the manuscript actually contains.
Introduction: Narrow from Broad to Specific
Think of the Introduction as a funnel that moves from the general to the specific:
- Opening context: Situate the reader in the broader research area
- Literature summary: Outline what the field already knows
- Knowledge gap: Identify what remains unanswered or contested
- Your study's role: Explain how this research fills the gap
- Aims and hypotheses: State clearly what you set out to investigate
The GMU Writing Center puts it well: "Begin by describing the problem or situation that motivates the research, move to discussing the current state of research in the field, then reveal a 'gap' or problem."
A frequent error here is previewing results or conclusions. Save those for their proper sections.
Methods: A Reproducible Recipe
The standard every Methods section must meet is reproducibility. A competent researcher working in the same domain should be able to replicate your study from this section alone.
What to cover:
- Study design with justification
- Participant or sample characteristics and selection strategy
- Materials and equipment, with enough specification for replication
- Procedures presented in the order they were carried out
- Data analysis approaches
- Ethical approvals and any relevant consent processes
Practical writing notes:
- Use past tense throughout ("We recruited participants...")
- Break complex workflows into subheadings
- Cite established protocols instead of reproducing them in full
- Include enough granular detail that gaps cannot be assumed away
Results: Data Without Editorial
The Results section is a reporting exercise, not an interpretive one. Keep your own opinions and speculation strictly out of this section.
What to include:
- Findings presented in a logical sequence, typically mirroring Methods
- Tables and figures as the primary vehicle for data
- Narrative text describing the most important trends and values
- Statistical test results with p-values, confidence intervals, and effect sizes
What to exclude:
- Any language that interprets or assigns significance beyond the numbers
- Repetition of every data point already visible in a table or figure
- Excessive raw data (move bulk data to an appendix)
Discussion: Putting Findings in Context
The Discussion section is where your results acquire meaning. According to Researcher.Life, a rigorous Discussion should address:
- Core findings: What are the headline results?
- Interpretation: Why do these results make sense, or why are they surprising?
- Connection to prior work: Where do your findings agree or diverge with the literature?
- Implications: What practical or theoretical doors does this open?
- Limitations: What aspects of the study design constrain interpretation?
- Future directions: What questions remain open for subsequent research?
A useful structural habit: open each paragraph with your interpretive claim, then support it with evidence from your own data and the wider literature.
Conclusion: The Final Word
Keep the Conclusion brief and grounded:
- Restate the principal findings in one or two sentences
- Articulate their broader significance
- Avoid introducing new data, claims, or citations
- Close by pointing toward future implications or open questions
Some journals fold Discussion and Conclusion into a single section. Always verify what the target journal expects before you draft.
Why Manuscripts Get Rejected
Most rejections trace back to a small set of recurring problems. Knowing them in advance gives you a concrete checklist to work against.
1. Scope Mismatch (Desk Rejection)
The problem: The manuscript does not align with the journal's stated aims and scope.
How to prevent it:
- Study the scope statement before submitting, not after
- Read through recent issues to gauge what the journal actually publishes
- Send a pre-submission inquiry for borderline cases
2. Insufficient Novelty
The problem: The paper does not add something demonstrably new to the field.
How to prevent it:
- State your novel contribution clearly and early in the Introduction
- Frame your work explicitly against what already exists in the literature
- Explain the specific advance your study represents
3. Methodological Weaknesses
The problem: Flawed design, underpowered sample sizes, or inappropriate statistical choices.
How to prevent it:
- Justify every major methodological decision
- Include statistical power calculations where applicable
- Bring in a statistician for studies with complex analytical demands
4. Unclear or Disorganized Writing
The problem: Reviewers cannot follow the argument, or the language creates unnecessary friction.
How to prevent it:
- Apply IMRaD structure consistently throughout
- Request collegial review before submitting
- Seek professional copyediting if English is not your primary language
5. Conclusions That Exceed the Data
The problem: The paper claims more than its findings can support.
How to prevent it:
- Tie every conclusion directly to a specific result
- Acknowledge limitations that affect how broadly findings can be generalized
- Use appropriately hedged language ("suggests," "indicates") rather than absolute assertions
6. Failure to Follow Journal Guidelines
The problem: Non-compliance with submission requirements triggers instant rejection at many journals.
How to prevent it:
- Read the author guidelines from start to finish before writing begins
- Use the journal's template when one is provided
- Cross-check reference style, word count, figure resolution, and any supplementary requirements
Pre-Submission Checklist
Run through this list before clicking submit:
Content and Argument
- Title is specific, concise, and accurate
- Abstract covers all sections within the word limit
- Introduction identifies the gap and states objectives
- Methods are sufficiently detailed for replication
- Results report data without interpretation
- Discussion interprets findings in relation to existing literature
- Conclusions are supported by the data presented
Figures and Tables
- Minimum 300 DPI resolution for all image files
- All axes, legends, and labels are clearly readable
- Visual style is consistent across the manuscript
- Every figure and table is cited in the text in order
For practical guidance on figure creation, visit our How to Make Figures for Nature/Science Journals guide.
Technical and Compliance
- Total word count falls within journal limits
- Reference style matches journal requirements exactly
- All citations are complete and accurate
- Ethical approval numbers are documented
- Conflicts of interest are declared
Final Review
- Proofread thoroughly for grammar and typographical errors
- Terminology is consistent from first page to last
- All numerical values and statistics are double-checked
- Author order and affiliations are confirmed
Principles for Effective Manuscript Figures
Strong figures are not decoration; they are your primary data communication tool.
Visual Design Fundamentals
- Maintain a consistent color palette across all figures in the manuscript
- Verify that text and data remain legible at the final print or screen size
- Add scale bars and units wherever spatial or quantitative context matters
- Avoid decorative 3D effects that distort perceived proportions
For color selection tailored to scientific work, see our Scientific Color Palette Guide.
Typography in Scientific Figures
- Prefer sans-serif typefaces such as Arial or Helvetica
- Target a minimum of 8pt for any text that will appear in the final publication
- Apply identical labeling conventions across all figures
For detailed font recommendations, visit our Best Fonts for Scientific Figures.
Frequently Asked Questions
What is the difference between a manuscript and a published paper?
A manuscript is the pre-publication document you submit to a journal for review. It becomes a paper or article once it has cleared peer review, been accepted, and gone through production. During that journey the manuscript typically goes through at least one round of revision based on reviewer feedback. Researchers sometimes use both terms interchangeably, but technically a manuscript refers to the version that exists before official publication.
How long should each section of an IMRaD paper be?
Proportions shift by discipline and journal, but a useful baseline for a standard 3,000 to 5,000 word research article is: Introduction around 15 to 20%, Methods around 20 to 30%, Results around 20 to 30%, and Discussion around 25 to 35%. The Abstract typically runs 250 to 300 words. Always defer to the target journal's author guidelines, since word limits and section expectations vary considerably across fields.
Should I write in first person or third person?
Modern scientific writing increasingly favors first person active voice, for example 'We analyzed the samples,' over passive constructions such as 'The samples were analyzed.' Flagship journals including Nature and Science now prefer active voice for clarity and directness. That said, certain disciplines and older journals still expect passive voice, so scan recent articles in your target journal and consult its author guidelines before committing to one register.
How do I respond to reviewer comments?
Prepare a point-by-point response letter that addresses every comment in order. For each point: quote the reviewer's comment verbatim, state whether you agree or disagree with it, describe the specific revision you made, and give the page and line number where the change appears. Maintain a professional and collegial tone throughout, including toward critical feedback. If you disagree with a suggestion, offer a clear, evidence-based counterargument rather than simply declining.
How many references should I include?
The appropriate number varies by paper type and discipline. Original research articles commonly cite 30 to 50 works, while comprehensive reviews may cite well over 100. Prioritize relevance, recency, and impact over sheer volume. Keep self-citation proportionate, generally below 10 to 15% of your total reference list, and favor primary sources over secondary summaries wherever possible.
What should I do if my paper is rejected?
Read the rejection letter and any reviewer comments carefully before reacting. If the issue is scope, identify a more suitable journal and submit there. If reviewers raised substantive methodological or content concerns, evaluate honestly whether revisions are feasible and worthwhile. A significant share of published papers were rejected at least once before finding a home, with some estimates putting that figure above 60%. Treat every rejection as a detailed critique that sharpens the manuscript before its next submission.
Can I submit to multiple journals simultaneously?
No. Simultaneous submission is considered a serious breach of publication ethics and can result in being blacklisted by journals. You must receive a final decision from one journal before submitting elsewhere. Because initial reviews often take two to four months, selecting the right journal at the outset matters. One permitted workaround is posting a preprint on a server like bioRxiv or SSRN while formal peer review proceeds, since most journals now allow this.
How long does the publication process take?
Timelines vary widely. Desk review and the decision to send a paper out for review typically takes one to four weeks. Peer review itself commonly runs one to three months. Revision and re-review can add another one to three months. Once accepted, online publication often follows within a few weeks, though print schedules may extend this. In total, the journey from first submission to publication commonly spans three months to over a year. Fast-track options exist for time-sensitive research, and preprint posting allows immediate dissemination while formal review continues.
Putting It All Together
A publishable research manuscript depends on two things working in tandem: a sound structure built on the IMRaD framework and a disciplined drafting process that starts with data and builds outward. Keeping these principles in view will serve you across every project:
- Build your figures and tables first so your data tells the story before a word is written
- Draft Methods early while every procedural detail is still fresh
- Defer the Introduction and Abstract until the full manuscript is drafted and your actual contribution is clear
- Understand what drives rejection and address those vulnerabilities proactively in every section
- Follow journal guidelines to the letter because formatting compliance is non-negotiable
Manuscript writing is inherently iterative. Multiple drafts, peer feedback from colleagues, and honest reflection on reviewer critiques are all part of the process. Rejection is a normal step on the path to publication, not a final verdict on your work.
Need to create professional figures for your manuscript? Try Figviz to generate publication-quality scientific diagrams from text descriptions. Perfect for researchers who want clear, accurate visuals without the steep learning curve of professional design software.
Author

Categories
More Posts

50+ Graphical Abstract Examples & Templates by Discipline (2026)
Browse graphical abstract examples from biology, chemistry, medicine, and more. Discover field-tested templates, layout analysis, and discipline-specific design tips for researchers.


Scientific Poster & Figure Fonts: The 2026 Typography Handbook
Choosing fonts for a scientific poster? Covers A0 and A1 font size charts, journal-specific requirements for Nature, Cell and ACS, plus the best free alternatives to Helvetica and Arial.


Research Data Visualization: Best Practices & Tools for Scientists (2026 Guide)
Master scientific data visualization with proven best practices. Learn to choose the right charts, use color effectively, and discover free tools for creating publication-ready figures.
