The publish dialog offers a short list: SCORM 1.2, SCORM 2004, maybe a few editions of 2004. Most people pick whichever was selected last time. That works until a report comes back looking wrong, and then the choice suddenly matters. SCORM 1.2 vs SCORM 2004: which version should you publish to, and what actually changes when you switch?

Publish to the version your LMS supports and reports most reliably, which is often SCORM 1.2 for simple courses and SCORM 2004 (3rd or 4th Edition) when you need separate completion and pass results or more saved progress. They differ in three practical ways. SCORM 2004 splits one status field into separate completion and success fields, it allows 64,000 characters of suspend data instead of 4,096, and it adds sequencing rules for multi-part courses. SCORM 1.2 is older, simpler and almost universally handled. The deciding factor is usually how your particular LMS stores and displays each one.

What Is SCORM, in Plain Terms?

SCORM, the Sharable Content Object Reference Model, is a set of rules that a course package and an LMS (learning management system) agree to follow. It is not a file format. It defines how a course is packaged so an LMS can import it, how the course talks to the LMS while it runs, and, in the 2004 version, how the LMS should move a learner between parts of a course.

SCORM 1.2 vs SCORM 2004 – Which Version Should You Publish To? - 3D Animation

The Advanced Distributed Learning (ADL) Initiative created SCORM. ADL's SCORM project page lists SCORM 1.2 from 2001 and SCORM 2004, whose 4th Edition from 2009 is the latest. The same page notes that ADL recommends xAPI and cmi5 for new development, while SCORM remains in very wide use. For most corporate training, SCORM is still the format the LMS expects.

Three terms are worth fixing before going further. A package is the zip file, with its manifest (imsmanifest.xml) at the root. A SCO, or Sharable Content Object, is a launchable unit inside the package that talks to the LMS. The run-time data model is the list of named fields, such as cmi.core.lesson_status, that the course reads and writes.

What Actually Changed Between SCORM 1.2 and SCORM 2004?

The data model, the API function names, the storage limits and the sequencing rules changed; the basic idea of a package talking to an LMS did not. These are the SCORM version differences a developer runs into most often:

AreaSCORM 1.2SCORM 2004 (3rd/4th Ed.)
StatusOne field: cmi.core.lesson_statusTwo fields: cmi.completion_status and cmi.success_status
Suspend data4,096 characters64,000 characters
Bookmarkcmi.core.lesson_locationcmi.location
ScoreRaw, min and maxRaw, min, max and scaled (-1 to 1)
API callsLMSInitialize, LMSSetValue, LMSCommit, LMSFinishInitialize, SetValue, Commit, Terminate
SequencingNo sequencing rules (only a basic prerequisites element that few LMSs honour)Rules for order, prerequisites and rollup

Two smaller changes trip up developers who hand-code against the API. First, the course finds the LMS differently: a SCORM 1.2 course looks for a JavaScript object named API in a parent or opener window, while a SCORM 2004 course looks for one named API_1484_11. A course that searches for the wrong name simply reports nothing.

Second, time is written differently. SCORM 1.2 expects session time as hours, minutes and seconds (for example 0000:12:30), while SCORM 2004 uses an ISO 8601 duration (for example PT12M30S). A malformed time value is enough to get a SetValue call rejected.

Exit values changed as well. SCORM 2004 adds "normal" to the list of values for cmi.exit, alongside "suspend", "logout" and a time-out value, which gives the LMS a clearer signal about whether the learner intends to come back.

The interactions model also grew richer in 2004, with better support for recording learner responses and descriptions of each question. That matters if anyone plans to analyze answers question by question in the LMS, which fewer teams do than intend to.

The side-by-side comparison of SCORM 1.2 and SCORM 2004 covers these differences in more detail, including rollup and the edition-by-edition changes.

One note on editions. SCORM 2004 went through four editions. The 2nd Edition capped suspend data at 4,000 characters; the 3rd and 4th raised it to 64,000. When a tool offers SCORM 2004 4th Edition, that is the current one, and most LMS platforms that support 2004 handle the 3rd and 4th Editions.

How Do the Status Fields Differ Between Versions?

SCORM 1.2 has one status field; SCORM 2004 has two. This is the difference with the most visible effect on reports.

In SCORM 1.2, cmi.core.lesson_status takes one of six values: passed, completed, failed, incomplete, browsed or not attempted. A course can only report one of them at a time. If it reports "passed", it is not separately reporting "completed", and the LMS has to infer the rest.

SCORM 2004 separates the two questions:

  • cmi.completion_status answers "did the learner finish?" with completed, incomplete, not attempted or unknown.
  • cmi.success_status answers "did the learner pass?" with passed, failed or unknown.

So a learner can be completed and failed at the same time in SCORM 2004, which is often exactly the truth. In SCORM 1.2, the course author has to choose which of those facts to send. Many tools ask this as a publish setting, something like "report status as passed/failed or completed/incomplete", and the right answer depends on what your LMS treats as done.

Does SCORM 2004 Store More Learner Progress?

Yes, a great deal more. Suspend data is the field a course uses to save its internal state between sessions, and the 3rd and 4th Editions of SCORM 2004 give it roughly fifteen times the room SCORM 1.2 does (64,000 characters against 4,096). The bookmark field is larger too: 255 characters for cmi.core.lesson_location in 1.2, and 1,000 for cmi.location in 2004.

For a short linear course, 4,096 characters is usually plenty. Problems appear in long courses with many tracked interactions, question banks, or a player that records the state of every slide. When saved state passes the limit, the LMS may refuse or truncate it, and learners return to find their progress gone. If that sounds familiar, the causes and fixes for suspend data and resume problems are set out step by step in a separate guide.

Switching to SCORM 2004 is one fix for a course that outgrows 1.2. It is not the only one, and it only helps if your LMS stores the full 64,000 characters, which is worth confirming rather than assuming.

What Is Sequencing, and Do You Need It?

Sequencing is a set of rules, written into the manifest, that tells the LMS how to move a learner between the SCOs in a package. It can require one module before another, hide a post-test until a pre-test is done, or decide how results from several SCOs roll up into one course result. Full sequencing exists only in SCORM 2004.

Most courses do not need it. A typical authored course is a single SCO with its own internal menu, and the authoring tool's player handles navigation. Sequencing becomes relevant when a package contains several separately launched modules and the LMS, not the course, must enforce the order between them.

Be cautious even then. LMS support for the full sequencing rules has historically varied, and complex rules are hard to test. If you need ordering between modules, check first whether your LMS can do it with its own course or curriculum settings.

Which SCORM Version Does Your LMS Handle Best?

Ask, then test. Supporting a version on paper and reporting it well are two different things. Put these questions to your LMS administrator or vendor:

  1. Which SCORM versions and editions does the LMS support?
  2. Which version do most existing courses on the system use?
  3. For SCORM 2004, do reports show completion and success separately, or merge them?
  4. Does the LMS store the full suspend data allowance for each version?
  5. Is there a sandbox or test area for uploading trial packages?

Then apply a simple rule. Choose SCORM 1.2 if the course is short, linear and judged on one result, and your LMS is happy with it. Choose SCORM 2004 4th Edition if you need completion and pass reported separately, the course saves a lot of state, or you need sequencing between SCOs. If your organization already has hundreds of 1.2 courses and reports built around them, staying consistent has real value too.

The cleanest way to decide is to publish the same course both ways and compare the LMS reports. AuthorLearn Studio can publish SCORM 1.2 and SCORM 2004 (3rd or 4th Edition) from the same course, which makes that comparison quick. You still need to check the results in your own LMS, because that is where the reporting behaviour lives.

SCORM 1.2 vs SCORM 2004 Comes Down to What Your LMS Reports Well

On paper, SCORM 2004 is the more capable standard. It separates completion from success, stores far more progress, and supports sequencing. SCORM 1.2 is simpler, older and almost universally supported. Neither is wrong.

The decision rule is practical: pick the version your LMS reports clearly for the kind of result you need, and confirm it with a test upload before you publish a whole library. A well-chosen version is one nobody has to think about again, which is the highest compliment a publish setting can earn.