7 best emission factor databases in 2026, compared

Every carbon platform quotes the size of its emission factor library, and the number tells you almost nothing about which of your suppliers to call. This page compares the seven databases those libraries are built from, on what each covers, how its factors were made, what it costs and, the question that matters for Scope 3, what it can and cannot tell you about a named supplier. DitchCarbon provides verified emissions data for over 2 million organisations, built on primary emissions data wherever it exists; a generic factor library drawn from five of these databases (ecoinvent, CEDA, EPA, DEFRA and EXIOBASE) is the fourth of four data layers, used only where the first three could not be found. Each entry below says whether it is in that library.

If you are choosing a database for your own Scope 1 and 2, or for the activity-based categories of Scope 3, the ranking below is the answer. If you are choosing a platform for Category 1 and 2, the ranking is the fallback that platform should be trying not to use, and the last two sections are the ones to read.

Compiled 23 September 2026 from each publisher's documentation.

What is an emission factor database?

An emission factor converts something you can count into tonnes of CO2 equivalent: a litre of diesel, a kWh of electricity in Poland, a euro spent on legal services. A database is a maintained set of those conversions with a stated method, boundary and year. Three families matter for corporate accounting.

Activity-based process factors measure a physical thing: a fuel, a material, a kilometre. DEFRA, EPA and ecoinvent are the reference sets. They are precise about the thing and silent about who supplied it.

Spend-based, environmentally extended input-output (EEIO) factors convert money spent in a sector into emissions, from national economic tables. EXIOBASE and CEDA are the reference sets. They cover every sector and every supplier at once, at the cost of giving every supplier in a sector the same intensity.

Aggregators collect the above under one licence and one API, with metadata on region, year and method. Climatiq is the reference example.

Under the GHG Protocol Scope 3 Standard, the hierarchy for Category 1 runs supplier-specific, hybrid, average-data, spend-based. Every database on this page sits in the bottom two rungs. That is not a criticism; it is what they are for.

The 7 best emission factor databases in 2026

Ranked by fit for corporate carbon accounting: breadth of activity and geography, transparency of method, update cadence and licence cost. What none of them can do for a named supplier is the subject of the section after the table.

1. ecoinvent

The reference life cycle inventory database: process-level factors for materials, energy, transport and manufacturing, with documented system models and unit processes you can open and trace. Swiss not-for-profit, licensed per user, updated on an annual cycle. Its depth is at the product and process level, which makes it the right base for a product carbon footprint and an expensive one for a spend ledger.

Who this is good for: product carbon footprints, LCA practitioners, and any team that needs to trace a factor back to the unit process it came from. In the DitchCarbon generic factor library: yes.

2. UK Government conversion factors (DEFRA, published by DESNZ)

Annual, free, and the default activity-based set for UK reporting: fuels, electricity, transport, water, waste, materials and hotel stays, with a full methodology paper each year; the 2026 set was published on 11 June 2026. Widely used outside the UK as the fallback for anything without a local factor. Weak on purchased goods and services, because the set is built around things a reporting company does itself rather than buys in.

Who this is good for: UK reporters, and anyone who wants a free, documented, annually refreshed set for Scope 1, Scope 2 and the activity-based parts of Scope 3. In the DitchCarbon generic factor library: yes.

3. US EPA GHG Emission Factors Hub and Supply Chain Factors

Two free sets. The Emission Factors Hub covers stationary and mobile combustion, refrigerants and US grid electricity by eGRID subregion. The Supply Chain GHG Emission Factors give spend-based factors for 1,016 US commodity categories at six-digit NAICS, built from the USEEIO model; version 1.4, released October 2025, uses 2023 emissions data and 2024 dollars. Together they cover most of a US company's Scope 1, 2 and spend-based Scope 3, for nothing.

Who this is good for: US companies that want Scope 1, 2 and a spend-based Scope 3 from one free public source. In the DitchCarbon generic factor library: yes.

4. EXIOBASE

The free multi-regional input-output database: 44 countries plus five rest-of-world regions and around 200 products per region, built by a consortium of European research institutes. The current release rests on supply and use tables to 2020, now-cast forward to 2022 with trade and macroeconomic data, so a factor applied to 2026 spend is carrying several years of change in grid mix, efficiency and price. Its breadth is unmatched among free sets; its resolution is a sector, not a company.

Who this is good for: multinational spend ledgers and financed emissions screens, where breadth across countries matters more than the age of the base year. In the DitchCarbon generic factor library: yes.

5. CEDA

The other EEIO reference set: around 60,000 spend-based factors across 400 industries and 148 countries, with mass-based factors for agricultural commodities alongside. Open CEDA is free under a CC BY-SA licence with attribution, and a premium edition sits above it. Finer resolution and fresher years than EXIOBASE, and it inherits the EEIO limit: every supplier in a category carries the same intensity.

Who this is good for: spend ledgers that need finer categories or a more recent base year than EXIOBASE offers. In the DitchCarbon generic factor library: yes.

6. Climatiq

Not a database so much as a warehouse: a licensed API over tens of thousands of factors from the sources above and many national ones, each carrying region, year, method, source and quality metadata, with a public explorer. For a team building its own calculator it removes the licensing and mapping work of assembling the sets individually. What arrives is still the source factor; the aggregation adds convenience, not specificity.

Who this is good for: teams building their own calculator who would rather license one API than assemble and maintain seven sources. In the DitchCarbon generic factor library: no; DitchCarbon licenses the source sets directly.

7. IEA emissions factors

Electricity and heat intensity by country, the reference for Scope 2 location-based factors outside the handful of countries publishing their own, updated annually and sold as a data product. Narrow by design: it settles one question, the grid, for everywhere.

Who this is good for: Scope 2 location-based reporting across many countries, and anyone modelling grid intensity over time. In the DitchCarbon generic factor library: no; grid factors come from the national sets above.

What a factor database cannot tell you

Which of your 40 professional services suppliers to call. A spend-based factor gives every one of them the same intensity per euro, so the ranking that comes out of a spend ledger is a ranking of your spend, not of their emissions. Two suppliers, same sector, same invoice total, identical footprint on paper, and one of them runs on renewable power and publishes a verified inventory while the other has never measured anything. The database cannot see the difference, because it was never looking at the supplier.

That is the reason platforms quote library size and the reason library size is the wrong question. DitchCarbon holds three layers before any generic factor is reached: corporate GHG inventories as reported by each organisation, supplier-specific emissions at spend, activity and product level, and the request that goes out, prepopulated, only where a primary figure would change the total. The generic library, drawn from ecoinvent, CEDA, EPA, DEFRA and EXIOBASE, closes the inventory where the first three could not be found, and the method behind every line is a filter, so the share of your baseline still sitting on a database factor is a number you can read rather than a footnote. Our primary versus secondary emissions data page sets out the difference in full.

Which database will your auditor accept?

Any of them, provided the choice is documented and applied consistently. The GHG Protocol names spend-based and average-data as valid Category 1 methods, and an auditor's question is not whether you used EXIOBASE but whether you can show which version, which year, which deflator, and why that factor for that line. Where the trouble starts is the year after: the standard expects you to move up the hierarchy as data allows, and an inventory that is still 100% on database factors in year three has a harder conversation ahead of it than one that shows the generic share falling.

For financed emissions the same point arrives as a score. PCAF data quality 4 and 5 are, by definition, the rungs where a factor database is doing the work, and moving a holding to DQ 2 or 3 means replacing the sector factor with the company's own reported figure. Our financed emissions software comparison ranks nine tools on that.

How do you check how much of your list sits on a database factor?

Ask. Send a sample of the organisations you buy from or invest in to any platform on your shortlist and ask the same three things: how many they hold a supplier-specific figure for, where each figure came from, and what happens to the rest. The honest answer to the third is the name of one of the seven databases above. What matters is how often that answer comes up.

DitchCarbon will show you what it already holds on your list, with the source and change history behind every figure. Numbers you can defend within 2 weeks.

Compiled 23 September 2026 from each publisher's documentation. Versions and licence terms change; confirm with the publisher before you build on one.