Evidence, disclosure and witness practice: cost and likely outcome in a Swedish commercial dispute depend chiefly on decisions taken before a claim is filed. Swedish courts do not order broad discovery; each side controls its own documents until a specific disclosure request succeeds, and outcome turns on whether witness accounts and technical records were preserved early enough to survive cross-examination.
Who this concerns
This matters to any commercial party that expects, or is already in, a dispute governed by Swedish procedural law, whether the forum is a Swedish district court or an arbitral tribunal seated in Sweden. It is relevant work within the disputes and arbitration practice at Lodline, and it applies equally to the party bringing a claim and the party defending one, because both sides face the same disclosure mechanics and the same rules on how testimony is taken.
The question comes up in three recurring situations. A counterparty has stopped performing and the company is assessing whether the paper trail supports a claim. A claim has already been served and the company needs to know what it can be compelled to hand over. Or a technical dispute, over quantum, specification or a construction defect, is turning into a battle of competing expert reports and the company wants to know how that battle is usually won or lost.
None of these situations is solved by general knowledge of Swedish civil procedure. They are solved by looking at what exists in a specific set of files, correspondence and contracts, and by knowing which of it can be forced out of the other side and which cannot.
What the law says
Under Swedish law as it currently stands, evidence in civil litigation is governed by the principle of free evaluation of evidence, known in Swedish as fri bevisprövning. There is no formal hierarchy between a signed contract, an email chain, a witness's oral account and an expert's written opinion, and no category of evidence is excluded purely because of its form. The court weighs everything presented and reaches its conclusion on the whole picture, which is precisely why the completeness of the record, not the eloquence of the advocate, tends to decide the outcome.
Disclosure works differently from what parties trained in common-law jurisdictions expect. Swedish procedure has no general discovery obligation. A party controls its own documents and is not obliged to volunteer material that helps the other side. What exists instead is a targeted disclosure duty, editionsplikt, under which a party can be ordered by the court to produce a specifically identified document that it is shown, or reasonably believed, to hold. The request has to describe the document closely enough that the court can rule on it; a request for "all correspondence relating to the project" is a different animal from a request for "the email of a named date attaching the revised specification," and only the second kind reliably succeeds.
Witness practice in Sweden is built around oral testimony given in person, under oath, subject to cross-examination by the other side. Written witness statements exist and are used to prepare a witness and to narrow what needs to be covered live, but a written statement rarely stands in for the witness appearing and answering questions. This is the point most often misunderstood by parties used to systems where a signed declaration can carry most of the evidentiary weight on its own.
How it works in practice
Preserving documents before a dispute exists
The single most consequential step happens before anyone has called it a dispute. Routine document retention and deletion policies keep running unless someone decides to suspend them, and a company that only starts preserving material once a claim letter arrives has typically already lost access to internal drafts, early correspondence and system logs that would have shown intent or knowledge at the relevant time. There is no penalty for documents destroyed under a genuine, unrelated retention schedule before a dispute was foreseeable, but the same destruction after a warning letter is read very differently by a court.
The scope of editionsplikt requests
A disclosure request under editionsplikt has to identify the document, not describe a category. Courts routinely reject requests that amount to asking the other side to search its own files for anything useful, and they routinely grant requests that name a specific email, a specific version of a contract, or a specific report referred to elsewhere in evidence already before the court. Building a request that will actually succeed means already knowing, from other correspondence or from a witness's account, that the document exists and roughly where it sits.
Witness statements versus oral testimony at trial
A written statement is a preparation tool. What decides the outcome is how that account performs live, under questioning designed to expose inconsistency between what was said then and what the documents show now. A witness whose account was recorded contemporaneously, close to the events described, tends to hold up. A witness reconstructing events years later from memory, with no note ever taken at the time, is the easiest target for the other side's cross-examination, regardless of how the case otherwise looks on paper.
Expert and technical evidence in commercial disputes
In disputes turning on quantum, technical specification or defect, the expert's report often becomes the actual battleground, and the underlying financial or technical data determines whether that report survives challenge. The pattern is visible in construction disputes over quantum, where the first financial figures produced usually set the frame the rest of the dispute has to argue within, for better or worse.
Cross-border evidence and witnesses
Where the counterparty, the relevant documents or a witness sit outside Sweden, the mechanics change. A foreign entity cannot simply be ordered to produce documents the way a Swedish party can; cross-border evidence usually has to go through mutual legal assistance channels or, within the EU, through instruments designed for that purpose, and both routes take materially longer than a domestic request. A witness resident abroad cannot be compelled to appear in a Swedish courtroom in the same way a domestic witness can, and testimony by video link or letters rogatory is weighed differently by the court than live testimony given in person. Where the dispute also involves tracing assets held by a debtor abroad, the evidentiary and enforcement questions start to overlap directly, which is covered separately in the context of tracing a debtor's assets abroad.
Arbitration versus court disclosure
Arbitral tribunals seated in Sweden are not bound by the same disclosure framework as the district courts. Document production in arbitration is governed by the parties' agreed rules and the tribunal's own discretion, which in practice tends to be somewhat broader than editionsplikt but far narrower than common-law discovery. A party choosing or already committed to arbitration is working under a materially different disclosure regime than the one described above for court proceedings.
What to check before the position is built
- Whether a litigation hold has actually been put in place, and when, relative to the first sign of dispute.
- Which documents are known to exist because they are referred to elsewhere, as opposed to documents merely assumed to exist.
- Whether any witness account has been recorded in writing close to the relevant events, or only exists as unrecorded memory.
- Whether any of the counterparty, key assets or a material witness sit outside Sweden, and what that means for how evidence can be obtained.
- Whether the dispute is heading to a Swedish court or to arbitration, since the disclosure mechanics differ.
- Whether expert or technical evidence is likely to be central, and whether the underlying data supporting it currently exists in a defensible form.
The numbers
Cost in this area is not driven by court fees; it is driven by volume and complexity. The variables that move cost most are the number of documents that need to be reviewed and, where relevant, translated; the number of witnesses who need to be prepared and examined; whether one or more experts are required and how contested their underlying data is; and how many rounds of disclosure requests and objections the dispute generates before the record is settled.
There is no fixed figure that applies across cases, and stating one would misrepresent how the process works. Timing is determined by the caseload of the specific court or tribunal seized of the matter and by how complete the submissions are when disclosure requests are made; a request supported by a precise description of the document sought is resolved faster than one the court has to narrow itself. The practical lesson is that cost and time are largely set by how much of the evidentiary groundwork was done before the request was ever filed, not by anything that happens afterwards.
Where it usually goes wrong
The approach above assumes a domestic Swedish court dispute with a cooperative, if reluctant, counterparty. It stops holding in several recognisable situations.
Where the seat is arbitration under institutional rules, disclosure is governed by the tribunal's discretion and the parties' agreement rather than by editionsplikt, and a party assuming court-style limits on production may find the tribunal willing to order considerably more, or considerably less, depending on how the rules and the tribunal are disposed. Anyone weighing that route specifically should look at how arbitration under SCC rules actually handles cost and disclosure before assuming court practice will simply transfer over.
Where a witness or a key document sits outside Sweden and outside the EU's cross-border assistance instruments, the practical ability to obtain either can collapse regardless of how strong the legal entitlement is on paper; a request that would succeed instantly against a domestic party can take many months, or fail outright, against a party or witness in a jurisdiction with no functioning mutual assistance channel.
Where documents were destroyed under a routine, pre-existing retention policy before any dispute was reasonably foreseeable, the destruction itself is not treated as concealment, and building an argument that it was tends to backfire with the court. The reverse also holds: destruction after a warning letter, even under a nominally routine policy, is read very differently.
Where the governing contract specifies a foreign law rather than Swedish law, the substantive rules on what has to be proved, and sometimes on how it has to be proved, can diverge from the Swedish procedural framework described here even where the forum remains Swedish, and the two should not be assumed to align automatically.
Does the position change when the counterparty or key assets are outside Sweden?
Yes. Cross-border document requests generally have to proceed through mutual legal assistance or, within the EU, through instruments designed for taking evidence abroad, and both take materially longer than a domestic request. A witness abroad cannot be compelled to attend in the same way, and the evidentiary weight given to remote or letters-rogatory testimony differs from testimony given live. Where the dispute also touches on locating a debtor's assets abroad, see tracing assets held by a debtor abroad for how the two issues interact.
What documentation is typically required in a technical or infrastructure-related dispute?
Technical disputes turn on specification documents, correspondence recording agreed changes, and any contemporaneous record of performance or non-performance against the relevant standard. The pattern of what a court or tribunal expects to see in this category, and what tends to be missing in practice, is set out for one recurring type of dispute in grid connection and capacity disputes, which illustrates the documentation gap that most often decides these cases either way.
Does the quality of the evidentiary record still matter once a judgment needs to be enforced abroad?
Yes, indirectly but significantly. A judgment built on a thin or contested evidentiary record is more exposed to challenge when a defendant resists recognition in a foreign jurisdiction, because objections to the underlying process are one of the few grounds available at that stage. How that plays out for one specific jurisdiction is addressed in recognition of a Swedish judgment in Italy, which shows how the evidentiary history of the original case can resurface at enforcement.
What to do next
This material goes as far as identifying what to preserve, what can realistically be requested, and where the usual approach stops applying to a specific dispute. It does not replace a review of the actual documents, the actual contract and the actual procedural posture of a specific matter, because that is where a disclosure request either succeeds or fails and where a witness account either holds up or does not.
Where the dispute is heading, or already committed, to institutional arbitration rather than court, the disclosure mechanics differ enough that they are addressed separately in arbitration under SCC rules. Where the question is instead whether the current record and the current documents support the position being taken, that assessment can be arranged directly through Lodline's contact channel.