The GHG Protocol scopes sort a company’s emissions into three tiers. Scope 1 is what you burn. Scope 2 is the energy you buy. Scope 3 is everything else your value chain does because you exist. Almost every reporting framework in use runs on them: in 2023, 97% of disclosing S&P 500 companies reported to CDP using the GHG Protocol.
Yet sustainability teams still disagree about where individual emissions belong — and the standard itself is now being revised, because the people governing it concluded the harder problem is no longer classification. It is data quality.
Scope 3 has fifteen defined categories, eight upstream and seven downstream. The five shown on each side are the ones that dominate most inventories.
This guide covers both: the boundaries, properly, and then the provenance question that matters more.
How much is at stake, in three numbers:
- 70–90% — the share of a typical corporate footprint sitting in Scope 3, outside the company’s own walls
- 97% — disclosing S&P 500 companies reporting to CDP using the GHG Protocol in 2023
- 10,000+ — companies now holding SBTi-validated targets, all of which mandate GHG Protocol methodology
The revision makes the shift toward data quality explicit. On 31 March 2026, after 42 sessions across more than twenty countries, a 65-member technical working group published its first public conclusions. The headline proposal was not a new category or a new calculation method. It was that companies should have to disclose what kind of data produced each number — which is, read plainly, an admission that the figures currently being reported are not reliably comparable.
Who should read this
- Carbon accounting practitioners
- ESG reporting managers
- Group financial controllers
- Sustainability consultants
- Internal auditors reviewing GHG data
- Anyone preparing for assurance
🔑 Key takeaways
- The GHG Protocol is effectively the global default. In 2023, 97% of disclosing S&P 500 companies used it to report to CDP, and over 22,100 companies disclosed through CDP in 2025.
- The organisational boundary comes before the scopes. Operational control, financial control or equity share — this choice determines what counts as yours, and changing it later makes your history non-comparable.
- Scope 2 always requires two numbers. Location-based and market-based, both disclosed. Reporting only one is the most common technical failure in a first inventory.
- Scope 3 is typically 70–90% of a total footprint and is where every boundary dispute lives. The 15 categories are exhaustive by design, not descriptive.
- The 2026 revision targets data quality, not boundaries. Proposals include mandatory disaggregation by data type, a 95% coverage threshold, a new Category 16, and disclosure of what has been third-party verified.
- Nothing has changed yet. The Phase 1 Progress Update is a working draft; public consultation is expected in the second half of 2026 and a final standard around late 2027. Build for the direction, report against the current standard.
On this page
- Why the GHG Protocol scopes exist
- The decision that comes first
- Scope 1 — direct emissions
- Scope 2 — purchased energy
- Scope 3 — the value chain
- The scopes in a real inventory
- Whose Scope 1 is your Scope 3?
- The one-question test
- The five boundary disputes
- Why provenance beats classification
- What the 2026 revision proposes
- What to do before the standard changes
- Common practitioner mistakes
- The bottom line
- Sources & verification
- Frequently asked questions
Why the GHG Protocol Scopes Exist at All
The GHG Protocol scopes are not an accounting convenience. They exist to solve a specific problem in climate mitigation: if every company simply added up all the emissions it touched, the same tonne of CO₂e would be counted many times over across an economy, and no one could tell who was responsible for reducing it.
The GHG Protocol Corporate Standard, first published in 2001, answers that by splitting emissions into three tiers of responsibility:
- Scope 1 — what you burn
- Scope 2 — what you buy as energy
- Scope 3 — everything else in your value chain that you influence but do not control
The split prevents double counting where it matters — inside one company’s total — while deliberately allowing it across companies. That overlap is not a flaw. It is what makes value chain pressure work at all.
That design choice is why the GHG Protocol scopes underpin almost every framework that follows. CSRD and the ESRS reference them. So do the ISSB standards, CDP, and the Science Based Targets initiative, which mandates GHG Protocol methodology outright. Change the Protocol and everything built on top of it moves.
The Decision That Comes First
Before any emission is assigned to a scope, an organisation has to decide what counts as itself. This is the organisational boundary, and getting it wrong invalidates everything downstream.
| Approach | What you account for | Typical user |
|---|---|---|
| Operational control | 100% of emissions from operations where you have authority to introduce and implement operating policies | The most common corporate choice; aligns with who can actually act on emissions |
| Financial control | 100% of emissions from operations where you direct financial and operating policies to gain economic benefit | Groups aligning the GHG boundary with the financial consolidation |
| Equity share | Emissions in proportion to your ownership percentage in each operation | Joint ventures, minority holdings, sectors with complex ownership |
The consequence practitioners underestimate. The boundary choice determines whether a joint-venture refinery sits in your Scope 1 or your Scope 3, and whether a leased building appears at all. Switching approach later re-bases your entire history — which is precisely the comparability problem a target relies on. Document the choice, the rationale, and any entities excluded, in the first year.
Scope 1 — Direct Emissions
Emissions from sources you own or control
Four source types, each measured differently:
- Stationary combustion — boilers, furnaces, generators. Fuel quantity × a fuel-specific emission factor.
- Mobile combustion — vehicles and off-road machinery you own or control.
- Fugitive emissions — unintended releases. Refrigerant leakage, SF₆ from switchgear, methane from process equipment.
- Process emissions — from chemical reactions, not combustion. Cement calcination is the classic case: CO₂ comes out of the limestone regardless of what heats it.
Scope 1 is usually the smallest number in a footprint and the one with the highest data confidence, because fuel purchases are metered and invoiced.
The trap: fugitive emissions get skipped because they feel immaterial. A supermarket’s refrigerant leakage can exceed all its combustion emissions combined, because refrigerant global warming potentials run into the thousands. Check the GWP set your framework requires — AR4, AR5 or AR6 — because the same leak produces different numbers under each.
Scope 2 — Purchased Energy
Indirect emissions from purchased electricity, steam, heat and cooling
Scope 2 is conceptually simple and procedurally the most frequently botched, because it requires two numbers, not one.
- Location-based — your consumption × the average factor of the grid you sit on. It answers: what did the physical grid emit to serve you?
- Market-based — the factor of the electricity you contractually bought: a PPA, certificates, a supplier tariff, or the residual mix. It answers: what did you choose to buy?
The Protocol requires both to be disclosed. They diverge sharply for any company buying renewable certificates, and the divergence is the point: location-based shows the physical reality of your grid, market-based shows the effect of your procurement decisions.
The trap: reporting only the market-based figure because it is lower. An assurance provider will ask for both, and a certificate claimed without checking vintage, geography and quality criteria will not survive review. Scope 2 is also under active revision — proposals would tighten the current annual matching of certificates toward more granular matching by location and time.
Scope 3 — The Value Chain
All other indirect emissions, in 15 defined categories
Scope 3 typically accounts for 70–90% of a corporate footprint and contains every genuinely hard question in carbon accounting. The 15 categories are exhaustive by construction — the standard intends that any value chain emission fits somewhere — which is why the boundaries between them generate so much argument.
| # | Category | What it captures |
|---|---|---|
| 1 | Purchased goods and services | Cradle-to-gate emissions of everything you buy; usually the largest single category |
| 2 | Capital goods | Embodied emissions of equipment and buildings, counted in full in the purchase year |
| 3 | Fuel- and energy-related activities | Upstream extraction and transport of fuels, plus grid transmission losses |
| 4 | Upstream transport & distribution | Inbound freight and logistics you pay for |
| 5 | Waste generated in operations | Treatment and disposal of your operational waste |
| 6 | Business travel | Employee travel in vehicles you do not own |
| 7 | Employee commuting | Travel between home and work, including remote-work energy where material |
| 8 | Upstream leased assets | Assets you lease in, where not already in Scope 1 or 2 |
| 9 | Downstream transport & distribution | Outbound logistics paid for by someone else |
| 10 | Processing of sold products | What downstream manufacturers do to your intermediate product |
| 11 | Use of sold products | Energy or fuel consumed by your product in use; dominant for energy-using goods |
| 12 | End-of-life treatment | Disposal of your products and packaging after use |
| 13 | Downstream leased assets | Assets you own and lease out to others |
| 14 | Franchises | Franchisee operations you do not control |
| 15 | Investments | Financed emissions; the core of PCAF accounting for financial institutions |
The GHG Protocol Scopes in a Real Inventory
Proportions on a diagram are abstract until you see them applied to a company you know. Apple publishes one of the most detailed corporate inventories in the world, which makes it a useful worked example — and an honest one, because the gaps in it are instructive too.
Three things in that chart are worth a practitioner’s attention.
The ratio is not unusual, it is typical. Apple’s operational emissions — everything it owns and every kilowatt-hour it buys — are roughly 7.7% of its footprint. The other 92.3% happens in factories, freight lanes and customers’ homes that Apple does not own. A company reporting Scope 1 and 2 only is reporting a rounding error and calling it a carbon footprint.
One category dominates. Purchased goods and services accounts for 54% of Apple’s Scope 3. This is the norm for anyone who manufactures a physical product, and it is why supplier engagement is the actual lever — your Scope 3 reduction is delivered by someone else reducing their Scope 1.
Even this level of disclosure has gaps. Apple provides category-level detail for 6 of the 15 Scope 3 categories. That is more than most companies publish, and it still leaves nine categories aggregated. Hold that against the disaggregation proposal below: the revision is not asking for something companies are already doing quietly. It is asking them to show work that is currently invisible.
Apple also reports a reduction of more than 60% against its 2015 baseline, explicitly excluding offsets. That distinction — reductions achieved versus reductions purchased — is one an assurance provider will test, and one a reader should look for in any climate claim.
Whose Scope 1 Is Your Scope 3?
The single most clarifying idea for a new practitioner is that scope is relative to the reporting entity, not intrinsic to the emission. The same tonne of CO₂e is Scope 1 for one company and Scope 3 for two others, simultaneously and correctly.
This also explains why supplier engagement functions at all: your Scope 3 reduction is achieved by someone else reducing their Scope 1.
The One-Question Test
Before the edge cases, the general rule. Almost every emission resolves into one of the GHG Protocol scopes through three questions asked in order, and getting the order right is most of the discipline.
One factory, every scope
Here is the same test applied to a single mid-size manufacturer. Nothing exotic — just the things a plant actually does in a week.
Two rows in that table are the ones people get wrong. The van and the haulier both move steel down the same road, but one is Scope 1 and one is Category 4, because you own one vehicle and merely paid for the other. And the customer collecting goods becomes Category 9 rather than Category 4 for the same reason in reverse — the invoice moved, so the category moved with it.
The Five Boundary Disputes
Five allocation questions generate most of the disagreement about the GHG Protocol scopes in practice, in review meetings and in assurance engagements. Each has a resolution rule that holds in the large majority of cases.
Where does this emission actually belong?
The rule in each case follows from a single principle: allocation follows control first, then who pays.
A leased warehouse — Scope 1 and 2, or Scope 3 Category 8?
The most frequent dispute, and it is decided by the organisational boundary rather than by the lease terms in isolation.
Rule: under operational control, if you set the operating policies for the site, its energy is your Scope 1 and 2. If you do not — a serviced office, a shared warehouse — it belongs in Category 8. Never both. Double counting between Scope 1/2 and Category 8 is one of the most common findings in a first assurance review.
Freight — Category 4 or Category 9?
Practitioners reach for “upstream means before us, downstream means after us”, which fails as soon as a customer arranges collection.
Rule: follow the invoice. Transport you pay for is Category 4, regardless of direction. Transport in your value chain that someone else pays for is Category 9. A retailer collecting from your factory is your Category 9, not your Category 4.
Employee travel — Category 6 or Category 7?
Blurred by hybrid working, field roles and travel that begins at home.
Rule: commuting is travel between home and a regular place of work — Category 7. Travel for a business purpose in a vehicle you do not own is Category 6. A field engineer driving from home directly to a client site is on business travel, not commuting. Home-working energy sits in Category 7 where material.
Category 3 — is this double counting Scope 1 and 2?
The category name says “not included in Scope 1 or Scope 2” and is still routinely misapplied.
Rule: Category 3 captures only the upstream portion — extracting, refining and transporting the fuel before you burn it (well-to-tank), plus transmission and distribution losses on electricity you purchased. The combustion itself stays in Scope 1; the grid electricity stays in Scope 2. If your Category 3 uses combustion factors, it is double counted.
Category 11 — do we have to model how customers use the product?
The category most often omitted, and for energy-using products almost always the largest.
Rule: if your product directly consumes energy or fuel in use, direct use-phase emissions are required, not optional. Model consumption per use, expected lifetime uses, and a regional grid or fuel factor — then disclose the assumptions, because they drive the result more than the arithmetic does.
Why Provenance Beats Classification
Return to the working group’s conclusion, because it reframes everything the GHG Protocol scopes are usually taught to do. A perfectly classified inventory built on spend-based estimates is less useful — and less defensible — than a roughly classified one built on supplier-specific data.
The reason is simple. Classification errors are visible and correctable. Provenance errors are invisible.
An auditor can see that freight was booked to Category 4 instead of 9, and ask you to move it. What an auditor cannot see — unless you disclose it — is that your Category 1 figure came from multiplying annual spend by an industry-average factor with no relationship to what you actually bought.
This is what the disaggregation proposal addresses. Requiring companies to publish emissions split by data type per category makes the difference between two reported numbers visible for the first time — and makes a company with worse data but better disclosure more credible than one with a confident single figure.
The practitioner test: for your largest Scope 3 category, can you state what proportion of the figure came from supplier-specific data, from activity data with secondary factors, and from spend-based estimation? If not, that is the gap the revision is designed to expose — and it is worth closing before disclosure becomes mandatory rather than after. Our free Carbon Accounting Methodology Selector works through the appropriate method category by category.
What the 2026 Revision Proposes
The Scope 3 Standard has not been meaningfully updated since 2011. Revision began in September 2024, and the Phase 1 Progress Update published on 31 March 2026 is the first public output.
| Proposal | What it would require | Who feels it first |
|---|---|---|
| Disaggregation by data type | Report Scope 3 emissions split into tiers by the type of data behind them, per category | Anyone relying heavily on spend-based estimation |
| 95% coverage threshold | Report at least 95% of required Scope 3 emissions to remain compliant with the standard | Companies currently reporting selected categories only |
| Verification disclosure | State whether the inventory is “Fully verified”, “Partially verified” or “Not verified” | Companies with partial or no third-party assurance |
| New Category 16 | A category for other value chain activities such as facilitated emissions and licensing | Financial institutions and licensors |
| Restricted allocation | Limits on allocating aggregated corporate-level emissions data down to products or activities | Companies using top-down allocation as a shortcut |
| Scope 2 granular matching | Under separate revision: matching certificates to consumption by location and time rather than annually | Heavy purchasers of renewable energy certificates |
Nothing here is in force. The Phase 1 Progress Update is a working draft approved for publication for transparency, not adopted as a standard. A full public consultation draft is expected in the second half of 2026, with a final revised standard targeted around late 2027. Phase 2 runs in parallel on category boundaries and the treatment of circularity and recycled materials. Report against the current standard; build toward the direction of travel.
Two structural developments matter alongside the text itself. In September 2025 the GHG Protocol and ISO announced a partnership to co-develop harmonised international standards for corporate GHG accounting, with ISO members joining the existing technical working groups. And in January 2026 the Protocol released a new Land Sector and Removals Standard, effective 2027, giving the first comprehensive guidance for land use, agriculture and carbon removals.
What to Do Before the Standard Changes
- Document your organisational boundary and stick to it. Write down the consolidation approach, the rationale, and every entity excluded. This is the first thing an assurance provider asks for and the last thing most teams have written down.
- Fix Scope 2 dual reporting now. If you publish only one figure, add the other this cycle. It is the cheapest credibility gain available and the granular-matching proposals will make certificate quality harder, not easier.
- Tag your Scope 3 data by provenance. For each category, record what share is supplier-specific, activity-based with secondary factors, or spend-based. You will need this if disaggregation is adopted, and you will learn something uncomfortable regardless.
- Measure your coverage against 95%. Identify which required categories you currently omit and what it would take to include them, even at low data quality. An estimated category disclosed as estimated beats a missing one.
- Resolve the five boundary disputes explicitly. Write the allocation rule your organisation applies for leased assets, freight direction, travel, Category 3 and use phase — then apply it consistently across years.
- Check your SBTi exposure. More than 10,000 companies now hold validated science-based targets, and the SBTi Corporate Net-Zero Standard v2 is being written to reference the revised Protocol. Any validated target will need the updated methodology at its five-year review.
Common Practitioner Mistakes
Boundary and classification
- Changing the consolidation approach mid-programme, making the baseline non-comparable
- Counting a leased asset in both Scope 1/2 and Category 8
- Applying combustion factors in Category 3 instead of well-to-tank factors
- Omitting Category 11 for an energy-using product because the modelling felt speculative
- Treating the 15 categories as a menu rather than an exhaustive set requiring a materiality judgement
Data and disclosure
- Publishing only the market-based Scope 2 figure
- Claiming certificates without checking vintage, geography and quality criteria
- Using the wrong GWP set for the framework you report under
- Presenting a spend-based estimate with the same confidence as measured data
- Skipping fugitive emissions on the assumption that refrigerants are immaterial
The Bottom Line
The GHG Protocol scopes are a mechanism for assigning responsibility across an economy without losing track of who can act. Scope 1 is what you burn, Scope 2 is what you buy as energy, Scope 3 is what your value chain does because you exist — and the same tonne being counted by three companies is the design working, not failing.
Learn the boundaries properly, because assurance providers test them and the five disputes above account for most of the findings. But do not mistake classification for rigour.
After forty-two sessions, the working group’s central conclusion was that the numbers need provenance, not better arithmetic. The most useful thing a practitioner can do before the revised standard lands is to be able to say, for every material category, where the figure came from — and to disclose it before anyone requires them to.
Sources & Verification
| Claim | Source & date |
|---|---|
| 97% of disclosing S&P 500 companies reported to CDP using the GHG Protocol (2023); 92% of Fortune 500 CDP respondents used it in 2016 | GHG Protocol, About Us and Standards pages — primary source, reviewed 28 July 2026 |
| Over 22,100 companies disclosed environmental data through CDP in 2025; approximately 45,000 suppliers requested to disclose via CDP’s Supply Chain programme | Industry analysis of GHG Protocol adoption, July 2026 — secondary reporting, not independently verified against CDP’s own release |
| Apple FY2025: Scope 1+2 operational emissions 1,265,400 tCO₂e; Scope 3 15,223,700 tCO₂e; purchased goods and services 54% of Scope 3; category detail published for 6 of 15 Scope 3 categories; upstream 67.4% / downstream 32.6% split | Apple corporate disclosures as aggregated by Tracenable and DitchCarbon, reviewed 28 July 2026. Figures are Apple’s own reported data; AiGreenTools has not audited the underlying inventory. |
| Apple reduced emissions more than 60% against a 2015 baseline, excluding offsets | Apple Newsroom, April 2025 — vendor-reported |
| Scope 3 Standard Phase 1 Progress Update published 31 March 2026; proposals on disaggregation by data type, verification disclosure and restricted allocation; explicitly a working draft, not a standard | GHG Protocol, Scope 3 Standard Revisions: Phase 1 Progress Update, 31 March 2026 |
| 95% Scope 3 coverage requirement and proposed new Category 16 covering facilitated emissions and licensing | ESG Today reporting on the Phase 1 Progress Update, April 2026 |
| 65-member Technical Working Group across 20+ countries, 42 sessions since September 2024; standard unchanged since 2011; public consultation draft expected H2 2026; final standard targeted late 2027 | Analyses of the Progress Update by Zevero (June 2026) and Seedling (June 2026), consistent across both |
| GHG Protocol and ISO partnership announced September 2025 to co-develop harmonised corporate GHG accounting standards; ISO members joining existing technical working groups | Seedling analysis, June 2026 |
| Land Sector and Removals Standard released January 2026, taking effect 2027 | Industry analysis of GHG Protocol updates, March 2026 |
| More than 10,000 companies hold validated SBTi targets; SBTi Corporate Net-Zero Standard v2 expected 2026–2028 and written to reference the revised Protocol | Zevero analysis, June 2026 |
| Scope 2 revision proposals toward granular matching of certificates by location and time, replacing annual matching | Zevero analysis of Scope 2 updates, July 2026 |
| Scope 3 typically represents 70–90% of a corporate footprint; three consolidation approaches; the 15 categories | GHG Protocol Corporate Value Chain (Scope 3) Standard and Corporate Standard — long-established, restated 2026 |
Every proposal described under the 2026 revision is a working draft subject to change through public consultation. This guide is an independent reference, not assurance or legal advice — confirm the current position with the GHG Protocol and your assurance provider before relying on it for a filing. Verified 28 July 2026.
Frequently Asked Questions
What is the difference between Scope 1, 2 and 3 emissions?
The GHG Protocol scopes split emissions by responsibility. Scope 1 is direct emissions from sources you own or control — combustion, company vehicles, refrigerant leakage, process reactions. Scope 2 is indirect emissions from purchased electricity, steam, heat and cooling. Scope 3 is everything else in your value chain, across 15 defined categories from purchased goods to investments. The split exists so that emissions can be aggregated without double counting within any one company’s total, while accepting deliberate double counting across companies — which is what allows value chain pressure to function.
Why does Scope 2 require two figures?
Because they answer different questions. Location-based uses the average emission factor of your physical grid and shows what the grid emitted to serve you. Market-based uses the factor of the electricity you contractually bought — a PPA, certificates, or the residual mix — and shows the effect of your procurement. The GHG Protocol requires both to be disclosed. Publishing only the market-based figure because it is lower is the most common technical failure in a first inventory, and an assurance provider will ask for both.
Is the GHG Protocol changing in 2026?
The GHG Protocol scopes themselves are unchanged, but the standard behind them is under revision. The Scope 3 Standard Phase 1 Progress Update of 31 March 2026 is a working draft from a 65-member technical working group, published for transparency rather than adopted. A full public consultation draft is expected in the second half of 2026 and a final standard around late 2027. The Corporate Standard and Scope 2 Guidance are being revised in parallel. Report against the current standard and build toward the proposals.
What is the proposed disaggregation requirement and why does it matter?
It would require companies to report Scope 3 emissions split into tiers according to the type of data behind them, category by category, rather than as one figure. It matters because it makes visible something currently hidden: whether a number came from supplier-specific data or from multiplying annual spend by an industry average. Two companies reporting the same total are not making the same claim, and disaggregation is what would let a reader tell them apart.
How do I decide whether a leased building is Scope 1 or Scope 3?
It follows from your organisational boundary. Under the operational control approach — the most common corporate choice — if you set the operating policies for the site, its energy is your Scope 1 and 2. If you do not, as with a serviced office, it belongs in Scope 3 Category 8. The one outcome that is always wrong is counting it in both, which is a frequent finding in first assurance reviews.
Which software helps with GHG Protocol accounting?
The right tool depends on where your programme is stuck rather than on the standard itself. If the constraint is producing a defensible number across many entities, carbon accounting platforms handle data collection and factor application. If it is deciding what to do with the number, decarbonisation planning tools model abatement cost. Our carbon accounting software ranking compares the field, and the free Methodology Selector identifies the appropriate calculation method for the data you already hold.
Where to Go Next
Put this into practice with the free Carbon Accounting Methodology Selector, which identifies the right GHG Protocol method for each Scope 3 category given the data you hold, or the ESRS Data Points Checklist if disclosure is the next obstacle. To choose a platform, see Best Carbon Accounting Software 2026, or SINAI Technologies vs Coolset for the measurement-versus-abatement distinction. For the regulations that make this mandatory, read What Is CSRD? and EU Taxonomy software. Browse every scored platform in the Carbon & Climate category. Primary source: the GHG Protocol.
