Recovering an unpaid B2B invoice in China usually runs through a formal demand, then negotiation, and only reaches litigation or arbitration once the debtor has ignored that notice. Which route fits your file depends on the invoice value, what the contract says about disputes, and the evidence you already hold.
This sequence fits an unpaid invoice where the buyer is a company registered in mainland China, the contract is silent on dispute resolution or names Chinese law, and you hold a signed order, a delivery record and the unpaid invoice itself. It works less well when the paper trail is thin, when the buyer disputes the goods or the service rather than simply not paying, or when the sum owed is small enough that the cost of building a file outweighs what you would recover.
Before committing time to any of this, check how much runway you actually have. A limitation period applies to commercial claims in China, and it is often shorter than creditors assume; the statute of limitations is the first thing to confirm against the contract date and the last acknowledgement of the debt, not the invoice date.
If the contract contains an arbitration clause naming a Chinese or foreign institution, the sequence below still applies up to the demand stage, but the escalation step changes. International arbitration for unpaid invoices is usually faster to a binding result than a mainland court, and an award is easier to enforce abroad than a Chinese court judgment.
Three things move the cost of this file more than anything else. The strength of the paper trail is the first: a file built on a signed contract and clear delivery evidence moves faster than one built on emails and assumptions. The forum is the second: arbitration under an institutional clause tends to be quicker and more predictable than mainland litigation, but it usually costs more upfront. The debtor's behaviour is the third.
A debtor who answers the demand, even to argue, is a different file from one who goes quiet. What happens when a Chinese debtor stops responding matters here, because silence usually means the next step is corporate research – checking the company's standing, its assets and whether it is still trading – using legal research and corporate intelligence from public and licensed sources, rather than another letter.
Distance adds its own cost. Instructing someone on the ground to review the file in Chinese, confirm the company's status and correspond with local counsel takes longer than a domestic collection, and every extra round of correspondence adds to the time before you reach a decision point.
At some stage – usually once the demand has gone unanswered and the debtor's status has been checked – you have enough information to decide whether to continue. The question is not whether you are owed the money; it is whether the debtor has assets worth pursuing and whether the claim's size still justifies the next stage's cost. This is the point to get a written view of the debtor's position before instructing litigation or arbitration, rather than after.
If the answer is yes, the practical next step is to instruct someone who runs China-facing recovery files as routine, not as an exception. Cross-border debt recovery in China at this stage means a demand backed by a credible escalation path, not a repeat of the letter you already sent.
Three conditions mean the file should close rather than escalate. First, the debtor company shows no active operations, no recoverable assets and no successor entity – a judgment against a shell is worthless whatever the court decides. Second, the amount owed is smaller than the realistic cost of litigation or arbitration once translation, local counsel and enforcement are counted in, and no settlement offer has closed that gap. Third, the limitation period has already run and no acknowledgement of the debt exists to restart it.
Before making that call, run the numbers rather than guessing them. A debt recovery cost calculator gives a rough comparison between what escalation is likely to cost and what the claim is actually worth, which is the comparison that should drive the stop decision, not sunk cost.
It depends on whether the debtor responds to the demand and whether the contract routes disputes to arbitration or to a mainland court. A debtor who negotiates can settle within weeks; one who ignores every notice and forces litigation or arbitration takes considerably longer, and enforcement can add further time after that.
You can file where your own courts have jurisdiction if the contract allows it, but a judgment from your home court is not automatically enforceable against a mainland Chinese debtor's assets. In most cases the realistic route is to pursue the claim, or enforce an award, in China itself.
Arbitration under an institutional clause is generally more predictable and often faster to a final decision than mainland litigation, and an award travels better internationally than a court judgment. It usually costs more to start, which is why the contract's dispute clause matters before a dispute exists, not after.
The invoice sat on your books while the shipment already left your yard, and every month it stays unpaid moves you closer to a limitation period you have not yet checked against the contract. A buyer who stops answering is not automatically insolvent, but the file only gets harder to build the longer it waits, and the balance sheet behind it does not stay still.