The Complete Guide to Scope 3 Under CSRD

I spent four months preparing my company's first voluntary TCFD disclosure in spreadsheets before I co-founded Carbonkindle. The Scope 3 section nearly broke us. The auditor rejected our figures the week before filing because we couldn't show data lineage from supplier invoices to the reported tonne. That experience convinced me that Scope 3 is not merely the hardest part of CSRD — it is the part that determines whether your entire disclosure holds up under assurance review.

Under ESRS E1, enterprises subject to mandatory CSRD reporting must disclose Scope 3 emissions across all material categories. Materiality here is double — financial materiality to the company and impact materiality on the climate system. Both tests must be applied before you can exclude a category. In our experience working with enterprise ESG teams preparing FY2025 and FY2026 filings, the single most common mistake is excluding categories without documented evidence of the materiality assessment process.

Understanding the 15 GHG Protocol Categories

The GHG Protocol Corporate Value Chain Standard defines 15 upstream and downstream categories of Scope 3 emissions. Not every category is material for every company, but all 15 must be considered in the assessment. Skipping the assessment is not the same as concluding immateriality — and the latter is what ESRS E1 requires.

Category	Type	Typical data source
Cat 1: Purchased goods & services	Upstream	Supplier primary data, spend-based proxy
Cat 2: Capital goods	Upstream	Asset purchase records, spend-based
Cat 3: Fuel- & energy-related	Upstream	Energy invoices, grid emission factor
Cat 4: Upstream transportation & distribution	Upstream	Logistics data, spend-based
Cat 5: Waste generated in operations	Upstream	Waste management records
Cat 6: Business travel	Upstream	Travel management system, expensed flights
Cat 7: Employee commuting	Upstream	Workforce survey, commute distance models
Cat 8: Upstream leased assets	Upstream	Lease records, building energy data
Cat 9: Downstream transport & distribution	Downstream	Customer delivery data, logistics provider
Cat 10: Processing of sold products	Downstream	Customer process data (intermediate goods only)
Cat 11: Use of sold products	Downstream	Product energy consumption specs, sales volume
Cat 12: End-of-life treatment	Downstream	Product materials data, disposal model
Cat 13: Downstream leased assets	Downstream	Tenant energy data (lessors only)
Cat 14: Franchises	Downstream	Franchise energy reporting (franchisors only)
Cat 15: Investments	Downstream	Financed emissions model (financial sector)

For most manufacturing and services enterprises, Categories 1, 3, 6, 7, and 11 will be the highest-volume categories. Category 15 (investments and financed emissions) is relevant to financial institutions and holding companies. The point is that you must assess each one, not assume.

What ESRS E1 Actually Demands

ESRS E1 is the climate-specific standard within the European Sustainability Reporting Standards, developed by EFRAG under the CSRD mandate. It requires disclosure of Scope 1, Scope 2, and total Scope 3 emissions by category — with clear identification of which categories are included, which are excluded, and why.

Specifically, ESRS E1-6 calls for gross Scope 3 emissions disclosed by upstream versus downstream categories, with the basis of consolidation (operational control or financial control), the emission factors used, and any recalculations from prior periods. The standard also requires narrative on the data quality for each included category — whether figures are based on primary supplier data, activity-based models, or spend-based proxies.

This is not a superficial checkbox exercise. Limited assurance under CSRD means a sustainability assurance provider will trace a sample of reported figures back to source data. If the lineage breaks — a supplier invoice that can't be matched to an emission factor calculation, for example — the figure is unacceptable and must be restated. Early CSRD filers have already discovered this: 1 in 3 Wave 1 CSRD filings contained Scope 3 restatements, according to assessments of early filings reviewed before formal publication of consolidated statistics.

Materiality Thresholds and Exclusion Rules

The GHG Protocol recommends excluding categories that collectively represent less than 5% of total Scope 3 emissions. ESRS E1 does not set a specific percentage threshold, but it does require that excluded categories be justified with a documented assessment.

In practice, we advise enterprise teams to run a preliminary spend-based estimate across all 15 categories before committing exclusions. Spend-based estimates are quick, defensible as a starting point, and often reveal that a category assumed to be immaterial actually represents 8–15% of total footprint. Cat 1 (purchased goods and services) is the classic case — because it aggregates all Tier 1 supplier emissions, it frequently exceeds 60% of a manufacturer's total Scope 3.

"The most expensive exclusion is the one you make before you run the numbers. A spend-based model takes a week to build and can prevent six months of restatement work." — Annika Lindqvist, CEO, Carbonkindle

Data Collection Strategies by Category Type

No enterprise collects primary data for all 15 categories in the first reporting year. The practical approach is a tiered data strategy: primary data for high-volume, high-auditability categories; activity-based models for moderate categories; spend-based proxies for lower-materiality categories where primary data is unavailable.

For upstream supply chain (Cat 1), the transition from spend-based to primary data typically takes two to three reporting cycles. The first year, you establish spend-based baselines using Exiobase or EEIO coefficients. In year two, you launch a supplier engagement program targeting vendors who collectively represent at least 80% of Scope 3 Cat 1 spend. In year three, you start substituting primary data for spend proxies and improving the weighted average data quality score for the category.

For business travel (Cat 6), most enterprises can reach activity-based accuracy using travel management system exports — distance by mode, IATA routing data, and IPCC aviation radiative forcing factors. This is often achievable in the first reporting year because the data already exists in procurement systems.

Employee commuting (Cat 7) is typically the most challenging. Workforce surveys are required because actual commute patterns are not observable in company data. Survey design matters: you need transport mode split, average distance, and remote-working days per week. Annualizing this requires some modeling, but the methodology is documented and reproducible.

Building a Defensible Audit Trail

The audit trail is what separates a disclosure that passes assurance from one that requires restatement. For every tonne of CO2e reported under Scope 3, there must be a traceable path: the raw activity figure (invoices, distance data, product quantities), the emission factor applied (source, version, year), the calculation formula, and the person and date of calculation.

When a supplier updates their emission intensity data after your initial filing — as frequently happens during the assurance window — you need to generate a restatement impact analysis immediately. This is operationally impossible in spreadsheets at enterprise scale. It requires version-controlled calculation infrastructure where recalculations propagate automatically.

Auditors from major assurance firms have told us directly that they reject disclosures where the calculation workbook cannot be re-run from original source inputs. The data lineage from source to reported figure must be unbroken. That is the baseline.

Practical Steps for Your Next Filing

1. Run a preliminary spend-based screen across all 15 categories using the prior year's procurement data and a standard EEIO coefficient database. This gives you a category-level footprint estimate within two to three weeks.
2. Document your materiality assessment before excluding any category. Record the assessment date, methodology, the estimated percentage of total Scope 3, and the basis for exclusion. Keep this in your CSRD documentation package.
3. Tier your data collection effort by category size. Allocate primary data collection resources to the categories that represent the top 80% of estimated Scope 3 emissions. Use proxies for the tail.
4. Stand up your audit trail infrastructure before data collection begins, not after. Retrofitting lineage documentation onto completed calculations is far more costly than building it in from the start.
5. Plan for restatements. Set a data-freeze date with a buffer before the disclosure filing deadline. Any supplier updates received after the freeze date go into next year's figures, not this year's restatement.

Scope 3 disclosure is a multi-year capability build, not a one-year compliance sprint. The enterprises that file credible disclosures in FY2025 and FY2026 are the ones that started treating Scope 3 as an accounting function — with the same rigor applied to financial close — rather than an estimation exercise. The CSRD assurance process will enforce that distinction whether companies are ready for it or not.