China Manufacturing Contracts: When One Agreement Is Not Enough
A U.S. company recently called one of our international dispute resolution lawyers after receiving a message from its Chinese factory. Future orders, the factory said, would require a 34% price increase, effective immediately. The factory had the client’s molds. It had the client’s designs. And it had a customer list the client never should have left exposed.
Two years earlier, trying to “keep things simple,” this American company had signed one contract with its Chinese manufacturer that combined tooling, NNN, and manufacturing terms into a single document. When we reviewed it, the problem was obvious: too many overlapping provisions, not enough clear remedies. It was hard to tell what the contract was supposed to do when the factory breached.
We counseled the client against suing the Chinese manufacturer to recover the molds. The contract was too muddled, the remedy too uncertain, and the business clock too unforgiving. We told them it would likely be cheaper, faster, and less risky to make new molds as quickly as possible.
The new molds took longer than expected, and the resulting production shutdown nearly bankrupted the company.
A cleaner, more focused Mold Ownership Agreement likely would have changed the factory’s calculation. A factory is far less likely to hold molds hostage when the contract gives the buyer a fast, obvious, and expensive remedy.
We have seen variations of this for years. The mistake is almost always the same: someone convinces the buyer that one document covering everything will be cleaner, faster, and cheaper than separate agreements or clearly separated contract sections. The problem is not the single document. The problem is the mush. Tooling rights, NNN obligations, quality standards, IP restrictions, and production terms get blended together until no one can quickly tell which breach triggers which remedy.
One contract is not always the problem. A single China Manufacturing Agreement can work if it is structured correctly. The problem is a contract that mixes different rights, remedies, and obligations so badly that no judge, arbitrator, or lawyer can quickly tell what should happen when the factory breaches.
Get the structure wrong, and the factory may have the leverage before the dispute even starts. Get it right, and the contract can do what it is supposed to do: make the factory think twice before breaching.
Separate Agreements or One Combined Contract?
If your Chinese factory breaches one obligation, can your lawyer quickly point to the relevant section, identify the breach, and explain the remedy without dragging in unrelated parts of the contract? If the answer is yes, one document may be enough. If the answer is no, the contract is probably too muddled to do its job.
A combined agreement can work when each part is clearly separated and enforceable on its own. It fails when NNN obligations, tooling rights, product development duties, quality standards, and production commitments are blended into one legal swamp.
That distinction matters because China manufacturing contract problems rarely stay theoretical. They show up when the factory refuses to return molds, uses your product design, sells around you, delivers defective goods, misses delivery windows, or demands a price increase after you have already handed over leverage.
Two Bad Assumptions That Sink Many Buyers
Foreign buyers often walk into China supplier relationships carrying two bad assumptions. The first is that the contract is paperwork: sign it, file it, forget it. The second is that the relationship matters more than the paper.
In China manufacturing, the contract often becomes the relationship when the relationship stops being easy. The factory that welcomed your first $200,000 order may feel very differently when you change suppliers, dispute an invoice, reject defective goods, or stop ordering altogether.
Chinese courts are generally not where you want to test vague contract language. You are far better off with a contract that clearly identifies the obligation, the breach, the remedy, and the basis for immediate relief. A Chinese-language contract that makes the violation plain gives the court or tribunal a practical path to act quickly. One that does not give the factory or the judge room to maneuver.
Chinese judges crave clarity. When they do not see it, they often lean hard on the parties to settle instead of issuing a ruling. We once had a case where the judge spent nearly three years urging settlement while the other side had no real interest in settling. By the time the court finally ruled, the case meant almost nothing to our client. A clearer contract likely would have produced a ruling much sooner.
The Four Agreements, and Why Each One Exists
The NNN: Why Your Western NDA Is a Paper Tiger
Before you send a single CAD file, bill of materials, sample, customer list, pricing sheet, or proprietary design to a Chinese factory you are evaluating, get a China NNN Agreement signed. If the information would hurt you in the hands of a competitor, it should not leave your office unprotected.
NNN stands for Non-Disclosure, Non-Use, and Non-Circumvention. Non-disclosure stops the factory from sharing your information with others, including a related factory down the road. Non-use stops the factory from using your information to make competing products for itself or others. Non-circumvention stops the factory from going around you to sell directly to your customers, whether through Amazon, Alibaba, direct outreach, or the buyers on your customer list. That is what lets you talk seriously with a potential supplier without handing over a roadmap to copy your product, build a competing version, or cut you out entirely.
A Western-style NDA does not solve this problem in China. Western NDAs are built around Western legal assumptions, Western litigation procedures, and Western remedies. They are usually not in Chinese, do not include enforceable liquidated damages calibrated for China, and do not address the two behaviors buyers most need to prevent: competitive use and circumvention. In some cases, they are worse than having no agreement at all because they give the buyer false confidence while leaving the factory room to maneuver.
Liquidated damages are what give a China NNN Agreement its teeth. They give a Chinese court a specific damages amount to apply without forcing the foreign buyer to prove every dollar of actual loss. The amount matters. Set it too high and the factory will argue it is an unenforceable penalty. Set it too low and it becomes a cost of doing business. Liquidated damages should usually be denominated in RMB for easier enforcement in China.
A well-crafted liquidated damages provision is often what makes a Chinese factory think twice before using or copying your product design.
We have drafted thousands of China NNN Agreements, and most are signed without meaningful pushback. Smaller buyers often worry they lack the leverage to get an NNN signed. That worry is usually overstated. A legitimate factory that wants your business will usually sign a properly drafted NNN.
One practical note: in China, a contract signed by an individual without the company’s official seal, commonly called the chop, can create serious problems. The chop is not magic by itself. You still need the correct contracting party, the correct Chinese company name, and a signature or chop process that matches how the company legally acts. But ignoring the chop issue is one of the easiest ways to hand the factory an avoidable defense and a terrific stalling tactic.
The Manufacturing Agreement: Where Most Disputes Are Won or Lost
Once you have picked your factory and are ready to place production orders, the Manufacturing Agreement becomes your rulebook. It is the document a Chinese judge or arbitrator will turn to first when something goes wrong.
Your China Manufacturing Agreement should cover pricing, quantities, delivery timelines, quality standards, inspection rights, payment terms, IP restrictions, tooling, exclusivity, force majeure, subcontracting, dispute resolution, and remedies.
A purchase order is not a substitute. A purchase order rarely covers the issues that matter when a dispute arises: quality standards, inspection rights, defective goods, late delivery, IP misuse, tooling, unauthorized subcontracting, and enforceable remedies.
Emails and WeChat messages can help prove what people said, but they are a poor substitute for a signed Chinese-language contract. WeChat records can disappear when a phone is lost or replaced, and emails can be difficult to authenticate in Chinese proceedings.
Vague language in your China Manufacturing Agreement is dangerous. A contract that says the product must be “good” or “high quality” is a gift to the factory. The agreement should specify objective, measurable standards. It should identify the applicable specifications, reference a physical golden sample held by the buyer, define inspection rights, and state what happens if the factory falls short.
Specifications should be attached to the contract or incorporated by clear reference. Too many buyers treat specifications as something floating in emails, purchase orders, or shared folders. That creates exactly the ambiguity your Chinese factory will exploit later.
The Manufacturing Agreement should also address whether the factory may sell an unbranded version of your product, what force majeure events excuse performance, how government-mandated shutdowns are handled, and whether unauthorized subcontracting is prohibited. Many tooling, quality, and IP problems begin when the factory quietly moves production or tooling to a related or lower-cost facility.
If you need exclusivity, define it carefully. A vague exclusivity clause can create more problems than it solves, especially if it does not specify products, territories, customers, duration, and remedies.
Governing law and dispute resolution should be chosen for enforcement, not comfort. Choosing New York or London law for a dispute you will need to resolve against a Chinese factory in China is usually a mistake. Most China manufacturing contracts should be written for enforcement in China, in Chinese, under Chinese law, with remedies a Chinese court or tribunal understands and will apply. If the contract is bilingual, the Chinese version should usually control unless there is a specific reason to do otherwise. The version that matters most is the one the Chinese court or tribunal will actually use.
Your choice of forum matters, and it should depend on the factory, its location, the contract value, the urgency, and the remedy needed. For urgent tooling, asset preservation, or evidence preservation issues, the dispute clause should be drafted with speed and enforceability in mind, not abstract neutrality.
If you sue in the United States against a Chinese factory with no U.S. assets, winning the case may be the easy part. Collecting on the judgment is usually the problem. Getting your tooling back or stopping IP infringement quickly will usually be impossible.
A factory-provided template is drafted to protect the factory, not the buyer. I have reviewed hundreds of factory-produced agreements, and I do not recall seeing one that adequately protected the foreign buyer. Roughly half were not even the right contract for the transaction. One common trick is to give the buyer a distribution agreement when a manufacturing agreement is clearly needed.
The Product Development Agreement: For Products That Do Not Exist Yet
If your product still has to be engineered, prototyped, tested, and accepted, you are not yet in production-contract territory. You are in development territory, and you need a different agreement to govern it.
If you are building a motherboard, a proprietary chemical blend, a custom mechanical part, a firmware-dependent consumer product, or anything else that needs real prototyping before it can be manufactured, you are in the high-stakes zone. You are not buying inventory yet. You are trying to create something that does not fully exist.
Most companies do not need a separate Product Development Agreement. Custom colors, logos, labels, packaging, and minor dimensional changes usually do not justify one. Genuine development does.
This is where buyers get ahead of themselves. A company shares a concept and preliminary specifications with a factory. The factory says it can make the product. The company signs a Manufacturing Agreement and pays a tooling fee before the product has been engineered, prototyped, tested, or accepted. Months later, the factory delivers samples that fall short. The factory says the samples match the specifications. The buyer disagrees.
Now the buyer is fighting about what was agreed, who owns the development work, and whether it is still bound by a Manufacturing Agreement it signed before knowing whether the factory could deliver. The factory has the buyer’s money, prototypes, tooling, and development history. The buyer has a launch deadline and shrinking leverage.
A Product Development Agreement forces the parties to separate development from production before the factory has your money, prototypes, and schedule in its hands. It should define milestones, deliverables, testing standards, rejection rights, payment triggers, ownership of development work, and what happens if development fails, including who bears development costs if the factory cannot meet the agreed specifications.
It should also separate what you brought into the relationship from what gets created during it. Your original drawings, formulas, brand assets, designs, and technical files need to stay yours. The improvements, modifications, tooling designs, test data, prototypes, firmware, and production know-how created along the way need to be assigned or allocated clearly. Without that language, ownership of factory-created improvements and modifications can become genuinely disputed under Chinese law.
Most importantly, the Product Development Agreement should make clear that no production obligation begins until the product has been approved. You never commit to buying something you have not accepted. Better yet, do not finalize the Manufacturing Agreement until the product has been developed and accepted.
Product Development Agreements take more patience to negotiate than NNN and Manufacturing Agreements. Chinese manufacturers sign NNN and Manufacturing Agreements regularly and often do so without much friction. Product Development Agreements are different because they limit the factory’s leverage by making clear that production is not guaranteed until the product works and has been accepted. Budget more time for the negotiation.
The Tooling Agreement: Preventing Mold Kidnapping
You can buy a mold, pay for the mold, have receipts for the mold, and still discover that “your” mold is sitting on a factory floor in Dongguan and not coming back.
That is the problem a Tooling Agreement, sometimes called a Mold Agreement, is meant to solve. In some cases, a self-contained tooling section inside the Manufacturing Agreement can do the job. Either way, the goal is the same: make clear who owns the molds, dies, jigs, fixtures, patterns, test equipment, and other production assets paid for or controlled by the buyer.
Paying for tooling does not always prove ownership clearly enough. In China, the fapiao can become important evidence of who paid for the tooling and who can claim ownership or control. The agreement should state who owns the tooling, who controls it, where it is located, how it may be used, and when it must be returned.
Most companies do not need a standalone Tooling Agreement. For many importers, a self-contained tooling section inside the Manufacturing Agreement is enough, provided it clearly addresses ownership, possession, permitted use, inspection, and return.
A standalone Tooling Agreement makes sense when losing access to the tooling would seriously disrupt the business. The issue is not only the dollar value of the molds. It is what happens if the factory refuses to release them. If replacing the tooling would shut down sales, delay a product launch, or cost you major customer accounts, a standalone Tooling Agreement deserves serious consideration.
Tooling disputes in China follow a predictable pattern. The buyer changes suppliers, payment gets disputed, quality collapses, or the factory realizes it has leverage. The factory, knowing the buyer cannot manufacture without its molds, refuses to return them. This is mold kidnapping, and it happens more often than importers expect.
When it happens, speed matters. Tooling can disappear, be modified, be damaged, or be moved to a subcontractor facility you have never visited and may not even know exists. Your goal is usually to get into court quickly and seek an order freezing, preserving, or returning the tooling before the dispute destroys your production schedule.
A standalone Tooling Agreement, or a clearly delineated tooling section within your Manufacturing Agreement, gives the court or tribunal what it needs. It establishes that the buyer owns or controls the tooling even though the factory physically possesses it. That claim should be supported by the fapiao, payment records, tooling inventory, photographs, identifying marks, serial numbers where available, weight records, and location records.
The agreement should also cover the practical details that determine whether the buyer can actually get the tooling back. It should restrict the factory from pledging, transferring, scrapping, lending, or using the tooling as leverage. It should limit use of the tooling to filling your orders only. It should address maintenance, wear and tear, moving costs, subcontractor use, notice requirements, and inspection rights.
Most importantly, it should have a clear process for demanding return of the tooling, a clear deadline for the factory to comply, and an appropriate liquidated damages provision if the factory refuses or delays. With this kind of agreement, many factories will begrudgingly return the tooling rather than fight a contract they are likely to lose on. Without it, the factory may see the tooling as leverage and treat it accordingly.
One tooling trap deserves special attention. Some factories offer to provide tooling at no upfront cost and recover the cost through per-unit pricing over a committed production volume. Without a clear written agreement establishing ownership and transfer rights from the start, the factory may argue that the tooling belongs to it until the volume commitment has been satisfied, or that you have no right to remove the tooling at all.
For companies with significant tooling exposure, a standalone agreement is worth the friction. Chinese manufacturers are less familiar with standalone Tooling Agreements than with NNN and Manufacturing Agreements, and they have little incentive to sign one enthusiastically. If the tooling is mission-critical, do not treat factory pushback as a reason to skip the agreement.
The Timeline of a Contract Disaster
China manufacturing contracts usually go wrong in sequence. The buyer does not lose leverage all at once. It leaks away one decision at a time.
The disaster often starts before the buyer even thinks it is in a dispute.
First, the buyer sends CAD files, technical drawings, samples, pricing, and customer information to three factories before anyone has signed an NNN Agreement. The factories are now holding the buyer’s product roadmap, and one of them may never become the buyer’s supplier.
Then, the buyer picks a factory and pays for prototypes without a Product Development Agreement. The factory now has the concept, the early design work, the prototype history, and the first real chance to argue later about who owns what was created.
Next, the buyer pays the tooling invoice before the tooling terms are signed and chopped. The mold is made in China, sits in the factory, and becomes the factory’s leverage the moment the relationship turns sour.
Finally, production begins on purchase orders, emails, WeChat messages, and a factory-friendly template that says almost nothing useful about inspection rights, defect remedies, subcontracting, IP misuse, delivery delays, or dispute resolution.
By the time the factory raises prices, delays shipment, sells around the buyer, or refuses to return the molds, the buyer is not negotiating from strength. It is trying to claw back leverage it already gave away.
The better sequence is not complicated. Sign the NNN before disclosure. Use a Product Development Agreement before real development begins. Put the Manufacturing Agreement in place before production orders. Lock down tooling before the factory makes, receives, or controls the molds, dies, jigs, fixtures, or other production assets.
The exact structure will vary. A simple branded product does not need the same contract architecture as a custom electronic device with tooling, firmware, and regulatory constraints. But the timing rarely changes. Protect information before disclosure. Define development before development. Define production before production. Lock down tooling before the factory controls it.
When One Document Can Actually Do the Job
A Manufacturing Agreement can include NNN provisions, tooling provisions, quality standards, inspection rights, IP restrictions, and dispute resolution provisions if each section is clearly separated and enforceable on its own.
We often incorporate NNN provisions and basic tooling provisions directly into Manufacturing Agreements. The same non-disclosure, non-use, and non-circumvention obligations can work inside a Manufacturing Agreement when they are separately titled, internally coherent, and enforceable on their own.
A Manufacturing Agreement with a clean NNN section and a clean tooling section is not a confused hybrid. It is a comprehensive document with clearly separated provisions a Chinese court or tribunal can read and apply without difficulty.
The test is practical. If the factory breaches one obligation, your lawyer should be able to turn to the relevant section, identify the breach, and explain the remedy without dragging in half the agreement.
If your lawyer starts stuttering and flipping through sixty pages to find the liquidated damages clause, you have already lost the first argument. Maybe not the case, but the clean, fast, cheap version of the case.
This is why combining NNN provisions inside a Manufacturing Agreement can be fine. Combining provisions is not the problem. Blurring obligations and remedies is. We often see this when the agreement is drafted by a lawyer who does not understand how Chinese courts, Chinese factories, and China manufacturing disputes actually work.
Red Flags Inside a Combined Agreement: A Lightning Round
Buried tooling terms are trouble. Tooling disputes need their own machinery: ownership, possession, location, permitted use, inspection, maintenance, and return provisions. They should not be tucked inside an IP paragraph that mostly talks about trade secrets and confidential information. When the factory refuses to return your molds, you do not want to be standing in court arguing that a general confidentiality clause somehow covers physical production assets sitting on the factory floor.
The phrase “quality products” should make you nervous unless it is backed by measurable standards. Quality without specifications, inspection rights, AQL standards where appropriate, acceptance procedures, and defect remedies is not quality protection. It is decoration. The contract should make clear what the product must be, how it will be inspected, who decides whether it passes, and what happens if it fails.
Generic IP protection language is another warning sign. A clause saying the supplier must “protect buyer’s IP” may sound good, but it is too soft unless it prohibits non-use, non-disclosure, non-circumvention, related-party misuse, subcontractor misuse, and competing sales. Broad IP language gives the factory room to argue. Specific prohibitions tied to specific remedies give you something to enforce.
Product development terms do not belong inside production commitments unless the contract is built very carefully. You should not be locked into buying a product before it has been engineered, tested, and accepted. A factory that controls the development process and already has a production commitment has leverage before you even know whether the product works.
Mold ownership language is not enough by itself. A contract that says you own the molds but does not identify the molds, their location, markings, fapiao, permitted uses, maintenance obligations, subcontractor rules, inspection rights, and return mechanism may leave you owning an argument rather than a mold. Ownership without a practical retrieval process often collapses when the factory refuses to cooperate.
A copied U.S. dispute resolution clause is usually a bad sign. The issue is not whether the clause looks normal to a U.S. lawyer. The issue is whether it gives you a practical path to relief against the factory, its assets, its tooling, and its conduct in China. If nobody can explain how the clause gets you a usable remedy in China, assume it will not.
An English-only contract governed by foreign law and aimed at a forum disconnected from the factory’s assets may protect you mostly on paper. Your factory knows that and will act accordingly. A contract that cannot be used quickly and effectively where the factory, molds, products, and records are located is often a contract that protects you only in theory.
And be suspicious of the agreement that looks impressive because it is long. Length can hide confusion. A contract can be expensive, dense, and still useless when the factory breaches. Bulk is not the goal. A useful contract tells the court or tribunal exactly what happened, exactly why it matters, and exactly what remedy should follow.
Why a Single Contract Cannot Do Four Jobs at Once Unless It Is Carefully Built
Think of a factory dispute as a messy divorce, but one where your ex-spouse has changed the locks on the house, kept the kids, and is currently liquidating your retirement account while you are still trying to negotiate who gets the toaster.
The factory has your molds. It has your product design. It knows your customers. It has your deposit. It has the production history. It may already have moved work to a subcontractor you never approved. Meanwhile, your contract treats all of these problems as if they were one big relationship issue rather than separate legal emergencies requiring separate remedies.
They are not the same problem. When a company comes to us after all of these problems have already merged, we often have to give them the hard truth: litigation may be slower, riskier, and less useful than rebuilding leverage elsewhere. That may mean moving production, recreating tooling, warning customers, changing product plans, or doing whatever else is needed to keep the business alive while the legal options are assessed.
An NNN dispute should be resolved on the terms of the NNN provisions, not tangled up with delivery timelines and quality standards from the Manufacturing Agreement. A tooling dispute should be resolved on the tooling provisions, not complicated by IP arguments from the NNN section. A product development dispute should be resolved under the Product Development Agreement, not under a Manufacturing Agreement for a product that was never accepted. Each agreement, or each clearly delineated section within an agreement, protects a distinct category of rights. Each must be enforceable without dragging in everything else.
When your spouse changes the locks, you do not want to spend three months arguing about retirement account valuation before you can get back into the house. When your factory refuses to return the molds, you do not want the court bogged down in a trade secret fight before it can decide who controls the tooling. What should have been a straightforward mold-return dispute can quickly become a contested interpretive exercise involving trade secret arguments, exclusivity claims, and counterclaims based on obligations the contract never clearly imposed.
When a contract is a mess, the opposing party can argue there is no clear obligation for the court or tribunal to enforce. It can inject IP arguments into a mold-return dispute and force the court to wade through trade secret questions when all you want is your tooling back. Delay almost always helps the Chinese factory because the factory already has the molds, product, information, or money. A clean contract denies the factory that tool.
A well-drafted, China-specific contract that is clear on obligations, breaches, and remedies is what works. Chinese courts are far more likely to enforce a contract quickly when it gives them a clean path to do so. We have moved from filing to asset freeze in a matter of weeks on tooling disputes where the contract was drafted to make the dispute simple. The factory had little room to argue, and the judge had little interpretive work to do. The tooling came back.
What This All Adds Up To
China manufacturing contracts are not judged by how tidy they look when signed. They are judged by whether they work when the factory breaches.
If your supplier relationship depends on confidential information, custom products, molds, tooling, or ongoing production, the contract structure matters as much as the contract language. The right agreement, signed at the right time, can prevent a dispute from becoming leverage in the factory’s hands. The wrong agreement can give the factory exactly the ambiguity it needs.
The buyer who waits until the factory has the molds, the designs, the samples, the customer list, the deposit, and the purchase orders is no longer building leverage. It is trying to recover leverage it already gave away. By that point, every option is worse than it would have been before disclosure, before tooling, before development, and before production began.
Build the contracts before the dispute starts. By the time the factory has your molds, designs, customer list, and production schedule, the leverage may already be gone.
Harris Sliwoski helps companies structure, draft, and negotiate China NNN Agreements, Manufacturing Agreements, Product Development Agreements, and Tooling Agreements, along with similar manufacturing contracts for Vietnam, Mexico, Thailand, and other key sourcing jurisdictions. We can help.
