SCORM 1.2 vs SCORM 2004: Which Should Training Providers Export?

SCORM 1.2 vs 2004 comes down to one trade. SCORM 1.2 is the most widely adopted version, while SCORM 2004 records more detail but only works as designed in an LMS that implements it well. For training providers who deliver to many client LMSs, the defensible default is to export SCORM 1.2.

Switch to SCORM 2004 (3rd Edition) course by course, when the client’s LMS is confirmed to support it and the course needs something 1.2 cannot do. A wrong choice tends to show up late, for example as a resume feature that misbehaves once bookmark data outgrows the limit.

This guide checks the numbers against the SCORM data model references and vendor documentation as of October 2026, and notes where sources disagree. It ends with a decision table and a pre-export checklist you can reuse for every client.

Key takeaways

• SCORM 1.2 has one status element, a 4,096-character bookmark limit and no sequencing. SCORM 2004 splits completion from success and adds sequencing, but the bookmark limit depends on the edition.
• Within SCORM 2004, the 2nd Edition allows only 4,000 characters of bookmark data, fewer than 1.2. The 3rd and 4th Editions allow 64,000.
• LMS support for SCORM 2004 is uneven. Moodle’s documentation says it is not supported, and TalentLMS says it supports only SCORM 1.2.
• Default to SCORM 1.2. Use SCORM 2004 3rd Edition when you can name the reason and have tested it in the client’s LMS or SCORM Cloud.
• When neither SCORM version fits, xAPI or cmi5 may, but cmi5 needs an LMS that implements a Learning Record Store.

What is the real difference in SCORM 1.2 vs 2004?

SCORM 2004 changed the data model, not just the version number. It splits the single 1.2 status into completion and success, raises the bookmark (suspend_data) limit from 4,096 to 64,000 characters in the 3rd and 4th Editions, and adds a sequencing and navigation specification that SCORM 1.2 does not have.

The table below uses the data model values from Rustici’s SCORM run-time reference and release dates from ADL’s SCORM page.

The 2nd Edition row matters. A package labeled “SCORM 2004” does not guarantee more bookmark room than 1.2 unless it is the 3rd or 4th Edition. ADL says it maintains resources for SCORM 1.2 and the 3rd and 4th Editions of SCORM 2004, so those are the practical export targets.

Which SCORM 2004 edition should you export, 3rd or 4th?

Export the 3rd Edition unless a client asks for the 4th or you need a 4th-only feature. ADL recommends the 3rd Edition for Department of Defense use, citing wider adoption and a more recently updated conformance test suite, and says it is about 99% compatible with the 4th.

The 4th Edition (2009) added four features: weighted progress-measure rollup, a “jump” navigation request, shared data between SCOs, and completion status on global objectives, according to Rustici’s 4th Edition announcement and its SCORM Golf examples. Many courses need none of these.

Sources differ here. iSpring’s guide advises the 4th Edition if the LMS supports it. That is reasonable when the LMS documents 4th Edition support. But Docebo’s help center lists only SCORM 1.2 and 2004 3rd Edition, so the 3rd Edition is the safer common denominator across clients.

Does SCORM 2004 track completion and pass or fail better than SCORM 1.2?

Yes, by design. SCORM 1.2 uses one element, lesson_status, to describe both whether the learner finished and whether they passed. SCORM 2004 reports completion_status and success_status separately, so a course can report “completed” and “failed” at the same time. That gives clients clearer reports on learners who finished but did not pass.

The gain comes with practical catches. Articulate’s documentation says that under SCORM 2004 learners must fully exit the course for completion data to show in LMS reports, and recommends adding an exit button. Its Rise 360 guidance says SCORM 2004 output can show duplicate Suspend, Continue and Close buttons unless you enable Hide LMS Interface.

Pro tip: Test The Fail Path. A quick test often checks only that a learner who passes shows as complete. Also run a learner who finishes but fails the quiz, then closes the window early, and look at what the LMS report shows. That is the case where the single 1.2 status and the split 2004 statuses can differ most.

When does the suspend_data limit push you to SCORM 2004?

Only when bookmark data really outgrows 4,096 characters. Suspend_data stores where the learner stopped and course state, so state-heavy or very long courses are the risk. For typical courses, SCORM 1.2 is enough, and the first step is to measure rather than assume.

Articulate’s support article says Storyline compresses suspend data, so exceeding the limit is unlikely, and that the limit depends on several factors you should test in your own LMS. If resume breaks, its suggested fixes are to turn off resume, reduce the slide count, or republish for SCORM 2004 3rd or 4th Edition. Trying the first two costs nothing and keeps the broadest LMS reach.

Which LMSs support SCORM 2004 and its sequencing?

It varies by LMS, edition and feature, so read the client’s documentation instead of assuming. Published help pages show real gaps: Moodle’s FAQ says SCORM 2004 is not supported, TalentLMS says it supports only SCORM 1.2, and Docebo lists SCORM 1.2 and 2004 3rd Edition with limits on sequencing.

Comparison posts from authoring-tool vendors often say most modern LMSs support both versions. One gives exact percentages, such as the share of LMSs supporting each version, but none of the posts I checked names a source for those figures, so this guide does not repeat them. Trust the LMS vendor’s own documentation for the LMS in front of you, and treat blog-level claims as unverified.

Rustici’s comparison page takes the same practical line from the content side: it calls SCORM 1.2 the most widely adopted standard and says creators who want the largest audience should be able to distribute in all three supported versions.

Pro tip: Ask For The Edition Name. “Do you support SCORM 2004?” can hide the difference between editions. Docebo, for example, names only the 3rd Edition. Ask which edition, whether sequencing is honored, and whether they have a sandbox tenant you can upload to before launch.

Is SCORM 2004 sequencing worth using for client courses?

Rarely. Sequencing controls what happens between SCOs, such as which one launches next and which navigation controls the learner sees. Because it governs movement between SCOs, it matters mainly in multi-SCO packages, and it depends on the LMS honoring the rules. If the LMS ignores them, the course can behave differently from your design.

Rustici’s sequencing guide says the specification was adopted from IMS Simple Sequencing and is “anything but simple.” It adds that with no sequencing rules the defaults should give an adequate experience, but some defaults can behave in counterintuitive ways. For most training providers, enforcing order through LMS prerequisites, or inside the course itself, is easier to test and to support.

SCORM 1.2 vs 2004: what should training providers export by default?

Export SCORM 1.2 by default, and move to SCORM 2004 3rd Edition course by course when you can name the reason and have tested it in the client’s LMS. This is a judgment drawn from the sources above, not a rule written into any standard.

Why 1.2: Rustici calls it the most widely adopted version, the LMS documentation above shows real gaps in SCORM 2004 support, and none of the sources I checked publishes verified LMS-share figures that would justify the opposite default. Use the table to decide per course.

What should you check before exporting a SCORM package for a client?

Confirm what the client’s LMS supports, decide what the course must track, then test the exact package you plan to send. The steps below move from the client’s LMS documentation, to the package itself, to a live test, and they catch most version problems before the client does.

1. Get the LMS name and version from the client, and find the vendor’s documented SCORM versions and editions yourself.
2. Ask whether the LMS honors SCORM 2004 sequencing, ignores it, or converts it to prerequisites.
3. List the tracking the contract needs: completion, score, pass or fail, time, interaction data.
4. Estimate bookmark size. Run the longest learner path and test resume at the very end.
5. Confirm your authoring tool exports the exact version you chose. Unzip the package and open imsmanifest.xml, which must sit at the package root, and read its schemaversion value (for example 1.2 or 2004 3rd Edition).
6. Upload to SCORM Cloud, which lists SCORM 1.2 and all three SCORM 2004 editions and lets you preview and debug. Its own page says it does not test for strict technical conformance, so treat it as a first pass.
7. Test in the client’s LMS or a sandbox tenant: complete, fail, quit early, relaunch, and check the learner report.
8. For SCORM 2004, test exit behavior and look for duplicate navigation buttons.
9. Keep the source files, and label each delivered zip with its version so you can re-export if the client changes LMS.

What if neither SCORM version fits: when to use xAPI or cmi5?

Use xAPI when you need to track learning outside an LMS, and cmi5 when you want xAPI tracking for content launched from an LMS. ADL recommends xAPI and cmi5 for new acquisitions and implementations, but both depend on infrastructure that the client’s LMS may not have.

xAPI records activity as statements, each with an actor, a verb and an object, sent to a Learning Record Store (LRS). An LRS can be part of an LMS or a standalone server. xAPI 1.0 was first released in 2013, according to ADL.

cmi5 is an xAPI profile for LMS-launched content. It adds rules for launch, authentication, session management, reporting and course structure. The cmi5 specification, whose first production release (Quartz) is dated June 1, 2016, requires the LMS to implement an LRS. It defines nine verbs: launched, initialized, completed, passed, failed, abandoned, waived, terminated and satisfied.

Check support before promising either one. xAPI.com says cmi5 adoption has been slowed by missing tooling and the lack of an authoritative source or governance for the specification. Articulate lists cmi5 as a Rise 360 export option and, from Storyline 360 build 3.39.21985.0, a Storyline publish option. If the client’s LMS says SCORM 1.2 only, as TalentLMS’s documentation does, neither is available there.

What should you do next?

Pick the default, then write a one-line intake question for every client: “Which LMS and version do you use, which SCORM versions and editions does it support, and can we test in a sandbox?” Export SCORM 1.2 unless the answer and the course’s needs justify 2004 3rd Edition. If you only have a finished package, our guide on how to edit a SCORM file without the source files and the SCORM rescue checklist for old client courses cover the next steps.

If you are converting existing material, the guides to a PDF to SCORM converter, an MP4 to SCORM converter and editable SCORM from PowerPoint show how each route handles tracking.

Disclosure: SimpliAuthor and SimpliTrain are made by the same company that publishes this site. SimpliAuthor’s SCORM converter page says it generates SCORM 1.2 or SCORM 2004 packages and lets you choose the version to match your LMS. It does not mention xAPI or cmi5. If you also run public or client-facing training programs, SimpliTrain has a page for training companies describing its platform.

Frequently asked questions