ACM Manuscript Formatting Guidelines
Table of Contents (8 sections)
ACM Quick Specification Matrix (2026 Author Protocol):
| Specification | Phase 1: Peer Review Draft | Phase 2: Camera-Ready (TAPS) |
|---|---|---|
| Layout & Columns | Single-column, standard A4 or US Letter | Two-column (sigconf/acmtog) or Single-column (acmsmall) |
| Line Spacing & Numbers | Double-spaced (2.0x), continuous line numbers | Single-spaced (1.0x), zero line numbers |
| Typography & Margins | 11–12pt Times New Roman, 1.0 in margins | 9–10pt body, 8pt references, strict template margins |
| Taxonomy & Metadata | CCS Concepts recommended, 5–10 keywords | CCS Concepts XML + display block strictly mandatory |
| Reference Block | Standard reference list in citation order | Canonical ACM Reference Format block required |
| Blind Review Protocol | Author masking & docProps metadata scrubbing | Full author identities, affiliations, and grant acknowledgments |
The Association for Computing Machinery (ACM) is the world’s largest computing society and the publisher of the most-cited venue portfolio in computer science. Its flagship journal, Communications of the ACM, has been in continuous publication since 1958. The ACM Transactions series — including ACM Transactions on Graphics (SIGGRAPH), ACM Transactions on Information Systems, and ACM Transactions on Software Engineering and Methodology — sets the standard for peer-reviewed CS research.
Beyond journals, ACM organises hundreds of annual conference proceedings, many of which are the primary publication venues in their subfields: SIGCHI (human-computer interaction), SIGMOD (database systems), ICSE (software engineering), CCS (computer and communications security), STOC (theory of computing), and many more. Papers published in these proceedings carry the same scholarly weight as journal articles in most CS tenure and promotion evaluations.
What makes ACM formatting uniquely challenging — and uniquely prone to submission errors — is that ACM has not one but two distinct, incompatible formatting phases that authors must navigate correctly. Getting the phase wrong is the most common single-source submission error in the entire ACM portfolio, and this guide will make sure you never make it.
2. The Two-Phase ACM Format — The Most Important Thing to Understand
Critical: ACM mandates two fundamentally different document formats depending on the stage of the submission process. Submitting the wrong format is the most common source of submission desk-returns and technical system errors in ACM venues.
Phase 1: Peer Review Submission
For initial peer review, ACM requires manuscripts to be submitted in a single-column, double-spaced format with continuous line numbers. This is the format reviewers actually read — dense two-column layouts at 9–10pt font are genuinely difficult to annotate and evaluate on screen. The peer-review format prioritises readability for reviewers over the compact, publication-efficient two-column layout.
Key characteristics of the peer-review format:
- Single column, occupying the full text width
- Double spacing (2.0x line spacing throughout)
- Continuous line numbers in the left margin
- Font size 11–12pt (Times New Roman or Liberation Serif)
- 1-inch margins on all sides
- CCS Concepts recommended (required in camera-ready)
- Often anonymized for double-blind review (venue-dependent)
Phase 2: Camera-Ready Submission
After a paper is accepted, ACM requires the author to reformat to the official camera-ready template. This is the format that will be published in the ACM Digital Library and proceedings PDF. The camera-ready format is two-column, single-spaced, uses a much smaller font (9–10pt), and must strictly conform to the ACM Word or LaTeX template.
ACM uses four primary document classes for camera-ready, each for a different venue type:
- sigconf: Standard conference proceedings (SIGCHI, SIGMOD, CCS, ICSE, etc.). Two-column, US Letter.
- acmsmall: ACM Transactions journals. Single-column, A4, wider margins.
- acmlarge: Some journals including JACM and TOSEM. Single-column, wide format.
- acmtog: ACM Transactions on Graphics (SIGGRAPH). Two-column, magazine-style layout.
The document class is specified in your submission’s Call for Papers. Always use the class listed there — do not guess based on what looks similar to a previous submission.
3. Core Document Formatting (Review Phase)
These settings apply universally to the peer-review submission format. They are the baseline requirements that every ACM reviewer expects to see, regardless of venue type.
- Font: Times New Roman or Liberation Serif, 11–12pt. Do not use sans-serif fonts (Arial, Helvetica) for the main text body of an ACM manuscript — the house style consistently uses serif typography.
- Line Spacing: Double-spaced (2.0x) throughout, including abstracts, section text, captions, and references. Single-spacing in any section outside tables is incorrect for the review phase.
- Line Numbers: Continuous line numbers are required for peer review submissions. Reviewers at ACM venues will reference line numbers in their written critiques — a submission without them is procedurally incomplete. Enable continuous (not restarting per page) line numbers in Word via Layout → Line Numbers → Continuous. See our guide on inserting continuous line numbers in Word.
- Margins: 1 inch (2.54 cm) on all sides. Standard US Letter or A4 page size depending on the specific conference — check the Call for Papers.
- Columns: Single column for peer review. Do not use Word’s column feature for the review submission phase.
- File Format: .docx (Microsoft Word). ACM provides an official Word submission template at authors.acm.org. Using the official template ensures proper style definitions that will process correctly through ACM’s TAPS production system after acceptance.
4. ACM-Specific Required Sections
Beyond standard academic manuscript sections, ACM imposes several unique content requirements that have no equivalent in other publishers. These are non-negotiable and will cause TAPS processing failures if absent from the camera-ready version, and editorial returns if absent from review submissions.
CCS Concepts
The ACM Computing Classification System (CCS) is a hierarchical taxonomy of computer science topics, maintained and updated by ACM. Every ACM paper must include a CCS Concepts block that categorizes the paper’s topics using this taxonomy. The CCS block serves as structured metadata that drives search and discoverability in the ACM Digital Library.
To generate your CCS Concepts:
- Visit dl.acm.org/ccs and browse the classification hierarchy.
- Select 2–5 categories that accurately describe your paper’s contribution. For each selected category, assign a significance level: Highlight (most central to the paper), General (related), or Related (peripherally relevant).
- The tool generates a formatted block in two parts: an XML/semantic annotation (for the LaTeX source or Word metadata) and a human-readable plain-text display for the paper body, formatted as:
CCS CONCEPTS • Computing methodologies → Machine learning; • Information systems → Information retrieval; • Human-centered computing → User interfaces.
Place the CCS Concepts block immediately after the Abstract and Keywords, before the Introduction. In the peer-review phase, CCS Concepts are strongly recommended; they are strictly required for camera-ready submissions.
Keywords
ACM requires 5–10 keywords that should ideally draw from the CCS taxonomy terms to ensure discoverability alignment. Keywords are placed immediately after the Abstract, before CCS Concepts, as a separate labelled block.
Abstract
ACM abstracts should be 150–250 words. Unlike many biomedical publishers, ACM does not require structured abstracts (Background / Methods / Results / Conclusions) for most venues — a concise, unstructured paragraph is standard. However, individual venue Call for Papers may specify a structured abstract format.
ACM Reference Format Block
Camera-ready submissions must include a formatted ACM Reference Format block immediately after the Author Keywords / CCS Concepts and before the Introduction. This block gives the canonical citation for the paper itself. For example:
Jane Smith and John Doe. 2026. Efficient Neural Architecture Search via Differentiable Proxies. In Proceedings of the 44th International Conference on Machine Learning (ICML ‘26). ACM, New York, NY, USA, 12 pages. https://doi.org/10.1145/xxxxxxx.xxxxxxx
The ACM Reference Format block is auto-populated in LaTeX when using the correct document class. For Word users, it must be typed manually based on the format provided in your acceptance notification.
Acknowledgments
The Acknowledgments section appears after Conclusions and before References. For double-blind review submissions, this section must be entirely omitted — it is one of the most common de-anonymization vectors. For camera-ready, include funding agency names with grant numbers, institutional affiliations, and any third-party tool acknowledgments required by the venue.
Availability / Artifact Appendix
Many ACM venues now feature an Artifact Evaluation (AE) track alongside the main review process. If your paper submits an artifact for evaluation, an Artifact Appendix must be appended to the paper. The appendix follows a standardized template (available from the venue’s AE committee) and describes the artifact’s purpose, access method, hardware/software requirements, and expected results for reviewers.
5. Review vs. Camera-Ready — A Comparison
The following table summarizes the key formatting differences between the peer-review submission phase and the camera-ready publication phase. Print or bookmark this table before starting your submission.
| Setting | Review Submission | Camera-Ready |
|---|---|---|
| Columns | Single | Two (sigconf/acmtog) or Single (acmsmall/acmlarge) |
| Line Spacing | Double (2.0x) | Single (1.0x) |
| Line Numbers | Yes — continuous required | No |
| Font Size | 11–12pt | 9–10pt (body), 8pt (references) |
| CCS Concepts | Recommended | Required |
| ACM Reference Format Block | Optional | Required |
| Anonymous | Often yes (double-blind venues) | No — full author details |
| Template | Flexible (Word or PDF) | Strict ACM template required |
| Page limit enforcement | Approximate (check CFP) | Strictly enforced by TAPS |
6. Double-Blind Review Requirements
The majority of ACM conference venues (and an increasing number of ACM Transactions journals) use double-blind peer review, where both authors and reviewers are anonymous to each other. If your target venue uses double-blind review, the following anonymization requirements apply to your initial submission. Failure to anonymize is one of the most common technical rejection causes at double-blind venues.
Removing Author Identity
- Title page: Remove all author names and institutional affiliations. Replace with a placeholder such as “[Author names and affiliations omitted for review]” or simply leave the author block empty per the template instructions.
- Acknowledgments: Omit the entire Acknowledgments section. Funding acknowledgments frequently name the institution or grant, revealing author identity.
- Running headers: If the template auto-inserts author names in running headers, update these to read “[Authors omitted]” or remove them.
Anonymizing Self-Citations
Self-citations must be anonymized in double-blind submissions. The correct approach is to cite your own prior work in the third person, as if it were someone else’s work, without omitting the citation entirely. Omitting citations to your prior work can create an incomplete or misleading literature review that damages your paper’s merit evaluation. The following patterns are correct:
- ✅ “Building on the approach described in [12], we extend...” (where [12] is your own prior work, listed normally in the reference list as [Author Omitted, 2024])
- ❌ “Building on our prior work [12], we extend...” (the possessive “our” reveals authorship)
- ❌ Removing the citation to your prior work entirely (creates an incomplete literature review)
Removing Metadata from the Word File
Microsoft Word embeds author identity information in the document metadata — specifically in the document properties (Author, Last Modified By) and in revision/comment histories. ACM’s TAPS system and many conference management platforms inspect this metadata. To scrub identifying metadata from a Word file:
- Open the file in Word and go to File → Info → Check for Issues → Inspect Document.
- In the Document Inspector dialog, check all relevant categories: Comments/Revisions/Annotations, Document Properties and Personal Information, Hidden Text.
- Click Inspect, then Remove All for each category that contains personal information.
- Save the file. Verify by re-opening Document Inspector to confirm no author metadata remains.
Important: Word’s Track Changes feature stores the name of the author who made each change. Accept all tracked changes before running Document Inspector, or the revision history will continue to expose author identity.
Our guide on cleaning Track Changes before journal submission walks through this process in detail.
7. 5 Common Submission Errors at ACM
The following five errors account for the majority of technical rejections, editorial desk returns, and TAPS processing failures in ACM submissions. Verify each item against your manuscript before submitting.
- Submitting two-column camera-ready format for initial peer review. This is the most common single error in ACM submissions. Authors who have previously submitted camera-ready papers, or who model their submission on a published ACM PDF, mistakenly use the two-column sigconf or acmtog layout for initial review. Reviewers must then read 9pt two-column text, which is significantly harder to evaluate and annotate. Many venue chairs will return two-column review submissions without review. Always check the venue’s Call for Papers explicitly for the review phase format.
- Missing or incorrectly generated CCS Concepts. CCS Concepts are required for camera-ready submissions and strongly recommended for review. Common errors include: using outdated CCS 2012 taxonomy terms (the current taxonomy is 2012, revised; verify at dl.acm.org/ccs), omitting significance levels (Highlight / General / Related), or formatting the CCS block as plain text without the required semantic annotation header. For LaTeX, the CCS XML block is embedded in the preamble. For Word, the plain-text display block must be placed correctly.
- Self-citations not anonymized for double-blind venues. This is a procedural disqualification at many ACM conferences. The first-person language “in our previous work” or “as we showed in [5]” reveals authorship identity. Program Chairs at major venues including SIGCHI, CCS, and ICSE have reported that failure to anonymize self-citations is among the top reasons for desk rejection without review at double-blind venues.
- Author metadata remaining in Word document properties. Even after removing the author block from the visible title page, Word documents store author identity in file properties and revision history. Submitting a Word file with author metadata intact to a double-blind venue is a deanonymization violation. Use File → Info → Inspect Document to scrub all personal information before submission.
- Using non-ACM-approved LaTeX packages causing TAPS failures. This specifically affects LaTeX users, but is worth noting for teams where Word and LaTeX authors collaborate. ACM’s TAPS (The ACM Publishing System) processes LaTeX source to generate the final PDF. TAPS rejects submissions that use non-approved packages or override ACM template styles. The approved package list is published at authors.acm.org. For Word users, submitting a non-template Word file can similarly cause TAPS style import failures. Always start from the official ACM Word template, not a blank document.
Before submitting to any ACM venue, verify that your tables follow the 3-line academic table standard and that continuous line numbers are enabled. If your paper uses Track Changes annotations from co-authors, follow our guide on cleaning Word Track Changes before submission to avoid metadata leaks. For a detailed breakdown of when to use single-column versus two-column format across publishers, see our companion guide on IEEE two-column vs single-column peer review.
Format Your ACM Manuscript for Peer Review
ScholarSpec formats your Word manuscript to ACM peer-review standards: single-column layout, double spacing, continuous line numbers, and 3-line academic tables — ready for initial submission.
Format for ACM Review in 1-Click8. Frequently Asked Questions
Should I submit my ACM paper in single-column or two-column format?
For initial peer review, submit in single-column, double-spaced format with continuous line numbers. Only convert to the two-column ACM template after acceptance, for the camera-ready version.
What are CCS Concepts and how do I generate them?
CCS Concepts are hierarchical taxonomy tags from the ACM Computing Classification System that categorize your paper’s topics. Generate them at dl.acm.org/ccs by browsing the taxonomy and selecting relevant categories and significance levels. The tool outputs a formatted XML block and plain-text display that must be included in your paper.
My ACM venue requires double-blind review. What do I need to anonymize?
Remove all author names, affiliations, and acknowledgments from the manuscript. Anonymize self-citations by referring to your own prior work in the third person. Remove identifying metadata from your Word file using File → Info → Check for Issues → Inspect Document.
Can I use Microsoft Word for ACM submissions?
Yes. ACM provides an official Word submission template on their website at authors.acm.org. For the camera-ready version, use the ACM Word template with the appropriate document class styles. Word submissions go through ACM’s TAPS processing system — follow the template closely to avoid TAPS errors.