Proving an Email Was Received: Who Has to Prove What Under German Law

A deadline expires, a termination notice is disputed, an offer is said to have been accepted in time. In all of these cases everything hinges on one question: did the email arrive – and when?

The answer comes in two parts. The first is a legal question: when is an email deemed received. The second is a question of evidence: who has to prove it, and with what. This article describes German law; other jurisdictions have their own rules.

When an email is received

A declaration of intent made to an absent person takes effect once it reaches the recipient (section 130(1) of the German Civil Code, BGB). It has reached them when it has entered their sphere of control in such a way that, under normal circumstances, they can take note of it.

For emails in business-to-business dealings, the German Federal Court of Justice has made this concrete: if the message is made available for retrieval on the recipient's mail server during normal business hours, it is received at that moment. Whether it is actually retrieved or read is irrelevant (BGH, judgment of 6 October 2022, VII ZR 895/21).

The recipient's mail server is therefore their sphere of control. That is the most important practical statement of the ruling, because it defines what the evidence has to show.

Who has to prove it

The sender. In full. And not only that they sent it, but that the message arrived.

The most common misconception here: that an entry in the sent folder or the absence of an error message is enough. Neither holds. Under German case law, a plain email without a delivery or read confirmation creates no prima facie evidence of receipt, even when no non-delivery notice came back (see, for example, Higher Regional Court Rostock, order of 3 April 2024).

The reason is technically understandable: a message can get stuck in a filter on the way, land in a quarantine folder or be moved by a rule without the sender ever finding out.

What works as evidence – and what does not

The screenshot of the sent folder. Proves sending, nothing more.

The read receipt. Proves a lot when it arrives. It only arrives if the recipient sends it, and they may decline. As plan A it is unsuitable, because the dispute is precisely the case in which nobody confirms.

The delivery receipt. Comes from the recipient's mail server and is therefore closer to what matters. It is not enabled everywhere, however, and many environments suppress it.

The sending system's log. This is where it becomes reliable – provided it records the right things and is protected against later change. A useful log records:

• which server the message was handed to, and when
• which response code that server returned, including the queue identifier it assigned itself
• how the connection was secured
• a hash of the bytes handed over, showing that exactly this content went out

That matches the Federal Court of Justice's criterion exactly: availability on the recipient's mail server. The receiving side's queue identifier is the most valuable part, because it does not come from you but from the recipient's system.

Why immutability makes the difference

A log that can be edited afterwards is open to challenge in a dispute – and the objection arrives reliably. So it is not enough to record the right details; they must also be provably unchanged.

The usual way to get there is chaining: each entry carries the hash of its predecessor, so any later change makes all following entries arithmetically invalid. If the head of that chain is additionally sealed at regular intervals and the seal stored immutably, rewriting the entire chain is ruled out as well.

Conbool SecureMail works on this principle: a chained log per tenant, a seal every fifteen minutes, and from it an evidence document as a PDF that includes the result of the chain check. If the message runs through the message portal, it also shows when it was opened and which attachment was downloaded.

What this is not either: a qualified timestamp from a trust service provider under eIDAS. The times come from the operation of the service. How a court weighs such evidence depends on the individual case.

Three practical recommendations

1. Do not rely on read receipts for time-critical declarations. They are an extra, not evidence.
2. Take shared mailboxes seriously. If time-critical mail arrives there, it counts as received even if nobody looks.
3. Have the evidence created before you need it. A delivery log cannot be produced retroactively – the receiving server's response exists only at the moment it is given.

This article provides a general overview and is not legal advice. For your specific case, please consult a lawyer.