Peppol BIS Billing 3.0 requires a fixed set of mandatory fields from the EN 16931 model, such as invoice number, issue date, seller and buyer identifiers, VAT data, tax categories, totals and the specification and process identifiers. Your software must populate each one in the required format, or the receiving access point can reject the document.
Key takeaways
- Peppol BIS Billing 3.0 invoices must carry specific header identifiers that tell the receiver which rule set applies.
- Seller and buyer endpoint identifiers must use a valid scheme and value that the Peppol network can resolve.
- Tax category codes, rates and totals must reconcile, so every VAT breakdown has to match the lines beneath it.
- Most first-time integration failures come from mapping errors in source data, not from the XML format itself.
What is Peppol BIS Billing 3.0?
Peppol BIS Billing 3.0 is a specification for exchanging electronic invoices over the Peppol network. It is built on the EN 16931 European semantic model, which names each business term and the rules that connect those terms. The Peppol billing documentation is the reference for every field discussed in this article, and you should check each field against it: Peppol BIS Billing 3.0 documentation.
Peppol moves documents through access points, which are gateways that route messages between senders and receivers. In most setups your software does not contact the buyer directly. It produces a structured XML invoice, submits it to an access point, and the access point validates it before delivery. An invoice that fails validation is often returned to the sender, so the error appears on your side at the moment of sending.
The European rules that drive this work are set by Directive 2014/55/EU, which concerns electronic invoicing in public procurement and asked contracting authorities to accept invoices that comply with the European standard: Directive 2014/55/EU on electronic invoicing in public procurement. This article is general information and not legal or tax advice; check the current text of the directive and your national rules with a qualified adviser.
Technical implementation of the specification is a separate task from interpreting the law. Engineers need the field list, the validation rules and the mapping logic. Rui Codex provides technical implementation support for teams building these integrations, including connections to the systems that hold invoice data, such as through its Peppol e-invoicing integration service.
Which invoice fields are mandatory?
Every invoice must contain the mandatory business terms defined in the specification, plus the header identifiers that tell the receiver which rule set applies. Some terms are conditional, so the presence of a field depends on the transaction, and you should read the cardinality column of the specification rather than assume.
The fields below are the ones that most often break a first integration. Business term numbers follow the EN 16931 convention (BT-x):
- BT-1 Invoice number: unique per seller and stable once issued. Do not reuse numbers after a cancellation.
- BT-2 Invoice issue date: the date of issue, in the format the specification requires.
- BT-3 Invoice type code: for example 380 for a commercial invoice and 381 for a credit note. A credit note sent with the wrong type code is one of the most common errors.
- BT-5 Document currency code: the currency in which the totals are expressed.
- BT-23 and BT-24 process and specification identifiers: the values that declare the Peppol BIS Billing 3.0 profile. These must match the profile the access point expects.
- BT-27 Seller name and BT-44 Buyer name: the legal names, written as registered.
- BT-109, BT-112 and BT-115 totals: the amount without VAT, the amount with VAT, and the amount due for payment.
Your mapping layer should treat these as hard requirements. A missing value should stop the send, not produce a placeholder that passes validation by accident.
How should you fill seller and buyer identifiers?
Fill the endpoint identifier with a scheme identifier and a value that the network can resolve, and fill the legal names and addresses exactly as they are registered. The endpoint is what lets the access point find where to deliver the document, so an error here stops delivery even when the XML is otherwise valid.
Seller and buyer endpoints each carry a scheme identifier that names the type of number used, such as a national business register number or a VAT number. The scheme values come from the Peppol code lists, which are part of the documentation linked above. Use the scheme that matches the identifier you actually hold. Mixing a VAT number with a register number under the wrong scheme is a common source of rejected documents.
Trade-offs appear when your customer master data is incomplete. Internal customer codes are convenient inside your ERP, but they are not endpoint identifiers. Storing the Peppol-ready identifier as a separate, validated field costs time at the start of a project, and it saves time on every invoice after that. Plan for a one-off data cleaning step before go-live.
How do VAT and tax category fields work?
Each VAT breakdown must carry a tax category code, a rate and a taxable amount that match the treatment applied to the lines, and the document totals must reconcile to those breakdowns. The category code describes the type of tax treatment, and choosing the right one is a tax decision that belongs to your accountant or tax adviser.
The software side of this is reconciliation. The taxable amount for each breakdown should equal the sum of the line amounts that share that category and rate, and the tax amount should follow from the rate. The document total should then equal the sum of the breakdowns plus any charges or allowances. Rounding rules matter here: if your system rounds per line but the total is rounded differently, the document can fail the sum check even though each line looks correct.
Keep rates and category codes in a reference table with effective dates. Hard-coding a rate in the invoice template is a maintenance risk, because rates and treatments change and old invoices must still reproduce their original values.
Hypothetical examples of field errors
Here are two hypothetical scenarios:
In the first scenario, a credit note is generated by the same template as a standard invoice, and the type code is left at the invoice value. The document passes the basic XML structure check, but the totals are read as a new charge rather than a reversal, which produces a mismatch between the sender's ledger and the receiver's records. The fix is a separate code path for credit notes, with its own type code and a required link to the original invoice number.
In the second scenario, a multi-line invoice is totalled by summing rounded line amounts, while the VAT breakdown is calculated from unrounded totals. Each line looks correct on its own, but the document-level sum check fails. The fix is to choose one rounding method in the specification of your own system and apply it at every level, then test the method with invoices that contain many lines of small amounts.
Where does this approach fail?
The approach fails when source data contains information the standard does not model, or when the mapping layer silently fills gaps. Field mapping works well for clean, consistent data, and it breaks down quickly when records were entered in free text, when one order spans several tax treatments, or when a legacy system uses codes that nobody documented.
There is a clear trade-off between validation early and validation late. Checking every field in your own system before submission gives fast feedback and a better audit trail, but it requires the team to maintain the rules as the specification changes. Relying only on the access point's validation messages is cheaper at the start, but the messages can be technical and the fix then lands on whoever reads the logs. Most teams benefit from both: a local pre-check for the common errors, and the access point as the final authority.
Integrations that connect an ERP to an e-commerce platform add another layer. Order data, returns and partial shipments all generate invoice events, and each event must map cleanly to a valid invoice or credit note. Teams planning this kind of work often benefit from reviewing the wider ERP and CRM integration approach before they design the invoice mapping.
Checklist before you send an invoice
- Confirm the process and specification identifiers match the Peppol BIS Billing 3.0 profile your access point expects.
- Check that the invoice number is unique and that credit notes use type code 381 and reference the original invoice.
- Verify that seller and buyer endpoints carry a valid scheme identifier and value.
- Confirm that every VAT breakdown reconciles with its lines and that the document totals match the breakdowns.
- Test rounding on invoices with many lines and small amounts.
- Log the validation response from the access point and route failures to a named owner.
If your team needs help building or testing these checks, you can request a project consultation to discuss the systems involved and the order in which to build them.
Frequently Asked Questions
Do I need to fill every field in the standard?
No, you need to fill every mandatory field and each conditional field that applies to your transaction. The specification marks which terms are required and under which conditions, so your mapping should follow those rules rather than filling everything. Leaving out a conditional field that does apply will cause validation to fail.
What is the difference between an access point and the Peppol network?
An access point is a gateway that sends and receives Peppol messages on behalf of a business, while the network is the set of access points and the directory that connects them. Your software usually connects to one access point and does not need to contact other participants directly. The access point handles validation and routing.
Why does my credit note fail when the invoice passes?
Credit notes use a different type code and usually require a reference to the original invoice. If the type code is left at the invoice value, or the reference is missing, the document can pass basic structure checks but fail the business rules. Give credit notes their own code path and test them separately.
Can I use one tax category for all my products?
Only if every product and service you invoice receives the same tax treatment, which is a question for your accountant or tax adviser. The software should support multiple categories and rates, so that the treatment chosen by your tax adviser can be recorded per line.
Where do I find the authoritative list of fields?
The authoritative list is the Peppol BIS Billing 3.0 documentation, which defines each term, its cardinality and the validation rules. The EN 16931 semantic model underlies those terms. Read the documentation for the version your access point supports, because rules can change between versions.
Conclusion
Peppol BIS Billing 3.0 is less about the XML file and more about the data behind it. The mandatory fields, endpoint identifiers and tax breakdowns each need a reliable source in your systems and a clear owner. Teams that treat validation as part of the send path, rather than an afterthought, find failures earlier and fix them faster. If you are planning an integration, email info@ruicodex.com to discuss your project and the invoice fields that matter for your systems.
Sources
- Peppol BIS Billing 3.0 documentation - OpenPEPPOL