Free PRISMA Flow Diagram Generator
Build a PRISMA 2020 diagram or scoping-review adaptation from your counts. Export PNG, SVG, editable PowerPoint or JSON. Free, no login.
Your review counts
For new reviews. Enter 0 for none; blank counts are not treated as zero. Records, reports and studies are counted separately.
Import a diagram JSON file
Drag a file here, or
Import a JSON file downloaded from this tool. Maximum 1 MB.
1. Identification
One per line: PubMed = 620. Up to 20 rows. Enter at least one database or register.
Optional. One per line: ClinicalTrials.gov = 12.
2. Screening and eligibility
One primary reason per excluded report. One row per reason: Wrong population = 60. Leave empty if none.
3. Included in the review
Enter independently: one study may have several reports.
After full-text exclusions; count shared reports once.
Live preview
Awaiting countsUnlimited diagrams, free without an account. Your data stays in this browser tab. Download JSON to keep your work before leaving.
From search to synthesis
Find papers, organise evidence and write your review with citations in AnswerThis. Download JSON first; this diagram does not transfer to an AnswerThis project.
How to make your PRISMA diagram
Choose systematic or scoping review. Enter each database and its record count, then the counts from your screening log. Use one line per database or exclusion reason, such as PubMed = 620 or Wrong population = 60.
Add the other-sources branch if you searched websites, organisations or citations. Enter the final included counts separately. Generate the diagram, check the preview and choose PNG, SVG, editable PowerPoint or JSON from Download.
This tool supports new reviews. Your counts stay in the browser; download JSON before leaving if you want to edit them later.
Check before publication
- Use actual counts from your search and screening logs. Blank is not zero. The generator does not search databases or screen papers.
- Count records, reports and studies separately. Multiple reports can describe one study; one report may describe more than one investigation.
- Report full texts you could not retrieve separately from reports excluded after eligibility assessment.
- Assign one primary exclusion reason per excluded report. Reconcile included reports shared between the database and other-source arms.
- Use PRISMA-ScR terminology for sources of evidence. A diagram is one reporting item, not a guarantee of compliance with the full checklist.
Worked example: 1,000 records become 30 reports about 24 studies
Suppose PubMed contributes 620 records and Scopus contributes 380. Identification totals 1,000 records. Remove 150 duplicate records, with zero automation or other removals, leaving 850 records to screen. Excluding 700 at title and abstract screening leaves 150 reports sought for retrieval. Ten reports cannot be retrieved, so 140 full reports are assessed for eligibility. Non-retrieval is a separate stage, not an eligibility exclusion reason.
At full-text assessment, exclude 60 reports for wrong population, 35 for wrong intervention and 15 for wrong study design. These primary reasons total 110, leaving 140 − 110 = 30 included reports. The review team identifies 24 distinct studies represented by those reports and enters 24 separately. The generator cannot infer 24 from 30: several reports may describe one investigation. Keep a report-to-study mapping in your review records to substantiate that distinction.
Reconcile overlaps and retain editable counts
If the databases arm includes 30 reports and an other-sources arm includes 8, with 3 reports shared by both, the unique total is 30 + 8 − 3 = 35. Record the overlap explicitly rather than adding both arms without adjustment. Reconcile each arm’s retrieval and eligibility counts first. Any records removed before retrieval need an explanation in the screening log; they should not simply disappear between identification and reports sought.
Download JSON before leaving so you can import the counts and resume editing. Use SVG for vector artwork, PNG for a flat image or PowerPoint when you need editable boxes, arrows and text. After changing counts, generate and check the exports again. Arithmetic checks do not confirm the screening decisions themselves. Counts and exports are processed locally, with no AI and no external metadata provider. Review the applicable reporting checklist alongside the diagram.
Based on PRISMA, with attribution
The diagram adapts the PRISMA 2020 flow diagram templates, available under CC BY 4.0. Source: Page MJ et al. BMJ 2021;372:n71. doi:10.1136/bmj.n71. Source-level counts and shared-report reconciliation are shown when applicable.
Scoping mode is a PRISMA 2020 layout adapted for PRISMA-ScR, not a separate official 2020-ScR template. Cite Tricco AC et al. Ann Intern Med. 2018;169:467–473. doi:10.7326/M18-0850. AnswerThis is not endorsed by PRISMA.
Frequently asked questions
Is the PRISMA diagram generator free?
Yes. Generate, edit, import and download as many diagrams as you need, without an account. Your counts are processed in your browser. Saving directly to an AnswerThis project is not available yet.
Does this search databases or screen papers?
No. Enter the actual counts and exclusion reasons from your review. The tool checks the arithmetic and draws the diagram; it does not search, deduplicate or screen papers.
Does it support PRISMA-ScR for scoping reviews?
Yes. Select PRISMA-ScR to use a PRISMA 2020 layout adapted for scoping reviews, with sources of evidence included rather than studies included. This is not a separate official PRISMA 2020-ScR template.
Can I edit my downloaded diagram?
PowerPoint downloads contain editable text, boxes and arrows. SVG is vector artwork and PNG is a flat image. Download JSON and import it here to restore your counts and regenerate every format.
Which PRISMA 2020 templates are supported?
New reviews using databases and registers, with an optional other-sources arm for websites, organisations or citation searching. Updated-review templates are not supported yet.
Does the diagram make my review PRISMA-compliant?
No. A flow diagram documents selection; you still need to complete the applicable PRISMA reporting checklist. Check your counts against your search and screening logs before publication.
What is the difference between records, reports and studies?
A record is a search entry. A report is a document about research. A study is the investigation itself and can have multiple reports. Enter the included study count separately; do not infer it from the number of reports.