Materiality Thresholds: How to Set Them in Your DMA (2026)

Every double materiality assessment (DMA) ends with the same question: from which score on is something material? The ESRS leave the answer to you. That freedom is useful, but it is also the point auditors probe first, because a threshold that moves by half a point can add or remove whole chapters of your sustainability statement.

In short: The ESRS do not prescribe materiality thresholds. Under the revised ESRS (2026), you set your own and justify them. In practice, companies score impacts on severity (scale, scope, irremediable character) and likelihood, and risks and opportunities on financial magnitude and likelihood, typically on a 1–5 scale with a threshold such as 3.0. Add override rules for severe cases (for potential human rights impacts, severity takes precedence over likelihood, ESRS 1 §40), anchor financial thresholds to your risk management, run a sensitivity check, and describe the result in ESRS 2 IRO-1. Auditors check that thresholds were defined upfront, justified, applied consistently and documented.

This article builds on our double materiality assessment example and uses the same scales and the same fictional manufacturer, PumpTec. If you are new to the two perspectives, read impact materiality vs. financial materiality first.

What is a materiality threshold?

A materiality threshold is the cut-off score above which an impact, risk or opportunity (IRO) counts as material. In a DMA you usually need two: one for impact materiality and one for financial materiality. A sustainability matter is material if at least one of its IROs passes either threshold. What exactly an IRO is, and why you should score IROs rather than topics, is explained in our article on IROs under the CSRD.

Not to be confused with audit materiality. In financial statement audits, "materiality threshold" means the amount above which a misstatement could influence users' decisions, often derived from profit before tax, revenue or total assets. A DMA threshold is different: it decides which sustainability matters you report on at all. The two meet only on the financial side, where it makes sense to reuse your company's financial and risk scales (see below).

What the ESRS require, and what they leave to you

Prescribed by the ESRS	Your choice (to be justified)
Impact criteria: severity (scale, scope, irremediable character) and, for potential impacts, likelihood	Scoring scales (1–5, 0–5, qualitative levels) and their level descriptions
Severity takes precedence over likelihood for potential human rights impacts (ESRS 1 §40)	How criteria are combined (average, maximum, product)
Financial criteria: likelihood and potential magnitude of financial effects (§50)	Numerical thresholds per dimension and any override rules
Time horizons: short (reporting period), medium (up to five years), long (more than five years)	Euro bands for financial magnitude, link to your ERM
Description of the process in ESRS 2 IRO-1, including when it was last updated	How IROs are aggregated to topics

The revised ESRS also clarify that quantitative scoring is not necessarily required and that obvious cases can be concluded top-down (§27). Thresholds therefore matter most for the topics you assess bottom-up. All changes are summarised in the DMA under the revised ESRS.

Impact materiality thresholds

Step 1: Fix the scoring model

On a 1–5 scale with written level descriptions, rate each impact on:

• Scale: how grave (or, for positive impacts, how beneficial) it is
• Scope: how widespread it is
• Irremediable character: how hard it is to reverse (negative impacts only)
• Likelihood: for potential impacts only

Step 2: Decide how to combine the criteria

This is where most of the hidden judgement sits. Three common options:

Method	How it works	Effect
Average	Severity = mean of scale, scope and irremediability; potential impacts: mean of severity and likelihood	Balanced and easy to explain; can dilute one extreme criterion
Maximum	Severity = highest of the three criteria	Conservative; tends to make more impacts material
Product	Score = severity × likelihood (1–25), threshold e.g. 9	Familiar from risk heat maps; punishes low likelihood strongly

PumpTec uses the average, a widespread and easy-to-explain choice. To see why the choice matters, take a hypothetical potential impact: groundwater contamination from a leak at the foundry, rated scale 4, scope 1, irremediable character 5, likelihood 2.

Method	Calculation	Score	Material at 3.0 (or 9)?
Average	Severity (4+1+5)/3 = 3.3; (3.3+2)/2	2.7	No
Maximum	Severity 5; (5+2)/2	3.5	Yes
Product	3.3 × 2	6.7 of 25	No
Average + override (irremediability 5 → material)	Override applies	–	Yes

Same facts, different outcomes. None of the methods is wrong, but you have to choose one before scoring and apply it to every IRO.

Step 3: Add override rules for severe cases

Averages are weak at protecting severe, unlikely impacts. Two override rules are worth writing into your methodology:

• Human rights override. For potential negative human rights impacts, severity takes precedence over likelihood (ESRS 1 §40). PumpTec applies this by using the severity score alone whenever it is 4.0 or higher. That is why forced or excessive overtime at Asian castings suppliers scores 4.7 and is material despite a likelihood of 2.
• Irreversibility override. Any negative impact with irremediable character 5 (cannot be reversed) is material, regardless of the combined score. This is a company choice, not an ESRS rule, but it is easy to defend.

Positive impacts are assessed on their own, without netting against negative impacts (§44). Some companies use the same threshold, others a separate one, for example to avoid reporting minor benefits. Either way, write it down.

Financial materiality thresholds

Link the magnitude scale to your risk management

Financial materiality is based on the likelihood and potential magnitude of financial effects (§50). The most defensible magnitude scale is the one your enterprise risk management (ERM) already uses. PumpTec (€540 million net turnover) calibrates it to EBIT impact:

Score	Magnitude (EBIT effect)	Likelihood within the time horizon
1	Less than €0.5m	Remote
2	€0.5m–2m	Unlikely
3	€2m–5m	Possible
4	€5m–15m	Likely
5	More than €15m	Virtually certain

The middle bands are illustrative; only the end points appear in our worked example. The financial score is the average of magnitude and likelihood, with the same 3.0 threshold.

Where the euro amounts come from

If your ERM has no usable bands, companies commonly derive them from a reference figure. These are common practice examples, not regulatory requirements:

Reference base	Typical for	Example anchor for score 3 (illustrative)
EBIT or profit before tax	Profitable industrial and service companies	Around 5% of planned EBIT
Revenue	Low-margin or loss-making companies	Around 0.5–1% of revenue
Equity or total assets	Capital-intensive companies, financial sector	Around 1% of equity

Whatever you choose, use the same base for risks and opportunities, and explain it in one sentence in your methodology.

Time horizons

Assess each risk and opportunity for the short, medium and long term and take the highest score, or assess it at the horizon where it is most likely to materialise. A carbon cost on steel inputs may be a 2 today and a 4 in the medium term. Discounting long-term effects is allowed if you do it consistently and say so.

Calibrating your thresholds

A threshold is only defensible if you have tested it. Four checks:

1. Dry run. Score 10 to 15 IROs from different topics before approving the methodology. If the results look absurd, fix the level descriptions, not the threshold.
2. Sensitivity check. Recalculate the result at neighbouring thresholds. For PumpTec (topic scores from the example):

Threshold	Material topics	What changes
2.5	5: E1, E5, S1, S2, G1	S2 becomes material on both dimensions
3.0	5: E1, E5, S1, S2, G1	Approved threshold
3.5	4: E1, E5, S1, S2	G1 drops out (financial score 3.0)
4.0	2: E1, S2	E5 and S1 drop out

The result is stable between 2.5 and 3.0, so the choice is robust. G1 Business conduct is the borderline topic and gets an explicit discussion in the calibration workshop. Also define whether a score equal to the threshold counts as material; PumpTec's rule says it does.

3. Peer comparison. Compare your material topics with the published sustainability statements of two or three sector peers. Differences are fine, but you should be able to explain them.
4. Consistency between topics. Do people rate "4" on a climate impact and "4" on a workforce impact with comparable severity in mind? A short calibration workshop with all raters catches most outliers.

Warning signs: if nearly all ten ESRS topics end up material, your scales are probably too generous or you are scoring topics instead of IROs. If only one or two are material for a company of PumpTec's size, check whether value chain impacts were scored at all. Our article on the materiality matrix shows how to visualise borderline topics.

Documentation and disclosure

What to write in ESRS 2 IRO-1

IRO-1 asks for a concise, company-specific description of your materiality assessment process, significant changes to it and when it was last updated. A description is only company-specific if it states your criteria and thresholds. A sample paragraph:

"Impacts were rated on a 1–5 scale for scale, scope, irremediable character and, for potential impacts, likelihood. For potential human rights impacts, severity took precedence over likelihood. Risks and opportunities were rated on financial magnitude, using the euro bands of our enterprise risk management, and likelihood over the short, medium and long term. An IRO was considered material at a score of 3.0 or higher in either dimension. The methodology was approved by the CFO on 15 March 2026 before scoring began."

Our ESRS 2 IRO-1 disclosure example contains a full model text.

Threshold policy template

Copy this table into your methodology document and fill in the right column:

Element	Your rule (example: PumpTec)
Scales	1–5 with written descriptions per level
Impact score	Average of severity and likelihood; actual impacts: severity only
Financial score	Average of magnitude (ERM euro bands) and likelihood
Thresholds	3.0 impact, 3.0 financial; equal to threshold = material
Override rules	Potential human rights impacts: severity alone if ≥ 4.0; irremediable character 5 = material
Positive impacts and opportunities	Same threshold
Aggregation to topics	Highest IRO score per dimension
Gross vs. net	Only implemented, effective measures reduce potential impacts (§43)
Approval	CFO, before scoring; date recorded
Changes	Only with documented reason; disclosed in IRO-1 if significant

Auditors will ask for exactly these points; our DMA audit checklist lists the evidence they typically request.

Common mistakes with materiality thresholds

1. Setting thresholds after scoring. If the approval date is later than the first scores, expect questions.
2. Tuning the threshold to reach a desired list. Fix the scales instead, and document why.
3. Averaging away severe impacts. Without override rules, a severe but unlikely impact disappears in the mean.
4. Financial scales that ignore company size. €1m is material for a mid-sized company and noise for a large group. Use your ERM bands.
5. Different raters, different scales. Without level descriptions and a calibration workshop, "3" means something different to every person.
6. No sensitivity check. If one topic flips at 2.9, you need to know before your auditor does.
7. Changing thresholds silently. A changed threshold is a change in methodology. Record it and, if significant, describe it in IRO-1.

Thresholds in Materiality Master

In Materiality Master, thresholds are part of the assessment settings. The default is 3 on a 0–5 scale, and you can set one threshold for all IROs or different thresholds per IRO type, for example for negative impacts and financial opportunities. When you change a threshold, the list of material topics and the materiality matrix update automatically, and so do the topics passed to the ESRS Datapoints Mapping Tool. That makes the sensitivity check above a matter of minutes. Every change to an IRO's scores is recorded in the audit trail. The help article on adjusting thresholds in a materiality assessment shows the steps.

Key takeaways

• The ESRS prescribe no thresholds; you set them and justify them.
• Fix scales, combination method and thresholds before scoring and have them approved.
• Use override rules so severe impacts are not averaged away; for potential human rights impacts, severity takes precedence (§40).
• Anchor financial thresholds to your ERM and treat percentages of EBIT, revenue or equity as company choices.
• Run a sensitivity check and discuss borderline topics explicitly.
• Describe criteria and thresholds in ESRS 2 IRO-1, with the approval date.

For the full process around the thresholds, see our CSRD double materiality assessment guide and the overview of the materiality assessment. Ready to set your own thresholds and see the effect immediately? Compare the plans and pricing for Materiality Master.