The 2026 ESRS List of Datapoints Is Here: How to Rebuild Your Disclosure Inventory Before 23 October

On 28 August 2026, the EFRAG Secretariat published the 2026 Draft List of ESRS Datapoints. It is a spreadsheet. It is also, right now, the most useful operational document in European sustainability reporting - and it is open for fatal-flaw review only until 23 October 2026.

If that sounds like a small thing, consider what the list actually does. The revised ESRS exist as a Delegated Regulation: legal text, organised by disclosure requirement, written to be read by lawyers. Your reporting system, by contrast, is organised by datapoint: a field, an owner, a source system, a validation rule, a control. The list of datapoints is the bridge between those two worlds. Without it, every company in Europe rebuilds the same mapping by hand, differently.

What EFRAG published

The 2026 Draft List reflects the ESRS as adopted by the European Commission in July 2026 through a Delegated Regulation, following EFRAG's simplification work in 2025. EFRAG describes it as effective from 2026 onward.

It was built using the same methodology as the IG 3 List of Datapoints published in 2024, with three usability improvements that matter more than they first appear:

• Datapoints are categorised using distinct data types. Narrative, monetary, quantitative, boolean, enumerated. This is the column that decides whether a field can be validated automatically and how it will eventually be tagged.
• Hyperlinks to the interactive ESRS in EFRAG's Knowledge Hub are embedded directly in the list, so a datapoint can be traced to its source paragraph without a separate document hunt.
• A separate version includes paragraph references back to the original 2023 ESRS, letting you see how individual disclosure requirements and datapoints evolved through the simplification.

That third item is the one to care about. It is, in effect, a crosswalk between the world your FY2024 and FY2025 reporting was built in and the world FY2027 will be reported in.

In parallel, EFRAG has prepared a Draft XBRL Taxonomy for ESRS aligned to the new draft list. It will be published separately for public consultation in due course, and both are covered in an August 2026 explanatory note.

Why this lands in a vacuum

EFRAG's implementation guidance documents - IG 1 on materiality assessment, IG 2 on value chain, IG 3 on datapoints - have not been updated for the revised standards. The ESRS Q&A Platform has been on hold since the end of 2025. For most of 2026, preparers working on the revised ESRS have had the legal text and very little else.

That is the context in which a draft spreadsheet becomes significant. EFRAG's SR TEG Chair, Chiara Del Prete, noted that EFRAG had received hundreds of requests for an updated datapoint list. The demand is not academic; it is people with a project plan and no map.

It also explains why the review window matters. Support material issued into a guidance vacuum tends to become the de facto standard, whether or not it was designed to carry that weight.

Support material is not law

Be precise about status, because your assurance provider will be.

The 2026 Draft List of ESRS Datapoints is EFRAG Secretariat support material, developed with input from the EFRAG Sustainability Reporting Board, SR TEG and the EFRAG Digital Reporting Consultative Forum. It is not endorsed by the European Commission and it does not create or remove obligations. The legal source of your disclosure requirements is the Delegated Regulation.

Practically, that means:

• You can use the list to structure your inventory, assign owners, scope system changes, and plan tagging work.
• You can cite it internally as the basis for a mapping decision.
• You cannot rely on it as authority if it and the legal text disagree. If you find a divergence, the text wins - and you have found a fatal flaw worth reporting.

There is a second timing caveat. The delegated act adopted on 3 July 2026 is still within the European Parliament and Council scrutiny period (two months, extendable by a further two). It applies once scrutiny concludes and it is published in the Official Journal. Commentators expect entry into force in the region of 10 November 2026, with mandatory application from FY2027 and optional early application for FY2026. Until then, everything below is preparation, not compliance.

The remapping job

Here is the sequence that works, assuming you already have a datapoint inventory from a previous reporting cycle.

1. Freeze your current inventory. Export what you actually collect today - field name, ESRS reference, data type, owner, source system, control. Not what your project plan says you collect. What the system contains.

2. Join on the paragraph-reference version. This is what the crosswalk version of the list is for. Match your existing references to the 2023 paragraph references, then follow them to the 2026 datapoints.

3. Classify every row. Five buckets, and be disciplined about the difference:

• Retained - same datapoint, same meaning. Leave alone.
• Reworded - same substance, different phrasing. Check whether the rewording narrows or widens scope before assuming it is cosmetic.
• Merged - two or more old datapoints now sit in one. These are the dangerous ones: they typically have two upstream owners who have never had to reconcile their definitions.
• Dropped - no longer required. The Commission reported a reduction of mandatory datapoints of roughly 61% where material, along with the removal of all voluntary disclosures and a stronger emphasis on the materiality-of-information principle. A large share of your inventory will land here.
• Newly conditional - still present, but gated differently. Materiality-of-information means fewer automatic obligations and more judgement you must be able to evidence.

4. Re-run the materiality gates. A dropped datapoint is a data-collection saving. A newly conditional one is a documentation obligation: you now need a reasoned, recorded basis for why you did or did not disclose it.

5. Re-point owners and controls. Every merged datapoint needs a single accountable owner. Every dropped datapoint needs removing from the collection template, not just from the report - otherwise you will spend FY2027 collecting data nobody uses and arguing about its quality.

6. Flag data type changes separately. A datapoint that survived but changed data type will not break anything in your FY2026 process. It will break validation and tagging later. Catch it now while it is a spreadsheet problem.

Why the data types column matters more than it looks

The Draft XBRL Taxonomy has been prepared in line with this list. That is the mechanism by which a column in a spreadsheet becomes a mandatory tagging rule in your filing pipeline: datapoint definitions and data types feed the taxonomy, the taxonomy feeds the digital tagging requirements, and the tagging requirements eventually land in your ESEF-equivalent filing process.

You do not need to solve tagging today. You do need to stop treating data types as metadata for the IT team, because a narrative field that should have been an enumerated one is a rework ticket you will pay for twice.

The fatal-flaw review: who should respond, and how

EFRAG has invited stakeholders to review the methodology used to compile the draft list and to report fatal flaws through an online survey by 23 October 2026. The final resource is expected by the end of 2026, once feedback has been considered.

Two things worth being clear about:

What counts as a fatal flaw. Mapping and methodology errors. A datapoint that exists in the Delegated Regulation but is missing from the list. A datapoint attributed to the wrong disclosure requirement. A data type that cannot represent the disclosure. A crosswalk reference that points at the wrong 2023 paragraph. An inconsistency between the list and the draft taxonomy.

What does not. Disagreeing with the Commission's policy choices. If you think a datapoint should not have been cut, the fatal-flaw survey is not the venue - that decision was made in the delegated act.

The people best placed to respond are the ones doing the remapping work described above, because they are the only ones who will encounter the errors. If your team hits a mapping problem in October and resolves it internally without reporting it, the same problem ships to every other preparer in the final version.

Between now and the end of 2026

• September: export and freeze your current datapoint inventory. Assign an owner for the remapping.
• Late September to mid-October: run the crosswalk and classify. Keep a log of every anomaly you find.
• By 23 October: submit fatal flaws from that log through EFRAG's survey.
• November: confirm the delegated act's publication in the Official Journal and its entry into force, then convert your draft mapping into a decision on FY2026 early application.
• December: update collection templates, retire dropped datapoints, and reconcile owners for merged ones before the FY2026 close cycle starts consuming everyone's attention.

None of this is glamorous work. But the companies that had a clean datapoint inventory going into the first CSRD cycle spent their time on disclosure quality, and the ones that did not spent it on reconciliation. The revised standards reset that inventory for everybody. The list EFRAG published in August is the cheapest available way to reset yours deliberately rather than by accident.

FAQ

Is the 2026 List of Datapoints legally binding? No. It is EFRAG Secretariat support material. Your legal obligations come from the Delegated Regulation containing the revised ESRS. Where the two appear to diverge, the legal text governs.

Do I have to use it? No. But there is no competing resource of equivalent coverage, and your auditors and software vendors are very likely working from it.

When will the final version be available? EFRAG expects to publish the final resource by the end of 2026, after considering fatal-flaw feedback submitted by 23 October 2026.

What happens to the 2024 IG 3 list? It remains the reference for the original 2023 ESRS. The 2026 list uses the same methodology and adds a version with paragraph references back to the 2023 standards, so the two can be read against each other.

Does this change my FY2026 report? Only if you early-adopt the revised ESRS. The revised standards are expected to apply mandatorily from FY2027, with optional early application for FY2026. The datapoint list is a preparation tool either way.

How does it relate to the XBRL taxonomy? The Draft XBRL Taxonomy for ESRS was prepared in line with the new draft list and will be published separately for public consultation. Datapoint definitions and data types in the list shape the tags in the taxonomy, which is why errors in the list are worth reporting now rather than after tagging becomes mandatory.