Information date: 31 August 2026 — China’s tax authority explains a permanent framework, effective from 2023, that allows eligible research and development expenses an additional deduction of 100 percent, with a corresponding 200 percent amortisation basis for qualifying intangible assets. These percentages describe tax treatment, not a cash grant equal to your research budget. For a foreign-invested business planning development work in China, the practical priority is a defensible link between eligible activity, cost records, accounting treatment and the taxpayer claiming the relief.
Separate the tax mechanism from a funding promise
The additional deduction operates within enterprise income tax rules. Whether an expenditure qualifies, how it is treated and whether your company can use the resulting deduction depend on the applicable conditions. You should not enter the headline percentage directly as cash income in a project model. The economic effect differs from a subsidy paid into a bank account.
The tax authority’s explanation also distinguishes expenses that do not form an intangible asset from those that do. Outsourced research introduces further rules, including the treatment of eligible commissioned expenditure. A common headline does not remove the need to check the arrangement, the recipient and the relevant documentation. Your tax team should confirm the detailed application rather than copying a percentage from a presentation.
This distinction matters before the research contract is signed. A company might have enough resources to begin development but depend on an assumed reimbursement to pay the next milestone. If that assumption is actually a tax deduction, the proposed cash schedule may not work. Funding approval should therefore stand on real equity, operating cash or committed financing, with tax effects modelled separately.
Connect the technical project to the accounting ledger
A technical narrative explains what uncertainty the project seeks to resolve, what work is performed and how progress is evidenced. The accounting ledger explains who incurred each cost, when it arose and how it was classified. Neither file is sufficient by itself. Generic descriptions such as innovation support make it difficult to determine which activities and expenses belong to the claim.
Consider a hypothetical team developing a new component while also performing routine customer customisation. Engineers may work on both during the same month. The useful control is a reasonable, supportable allocation based on actual work, not an automatic decision to label every engineering salary as eligible R&D. The appropriate allocation and eligible categories require tax review.
Three effects deserve management attention. Poor separation can increase tax risk; late reconstruction of records consumes engineering and finance time; and overstated relief can distort the decision to continue a project. A shared project code helps connect technical evidence, purchasing and payroll without allowing any one code to determine eligibility automatically.
For outsourced work, the contract should describe deliverables, rights, responsibilities and supporting information. An invoice saying research services does not, on its own, establish the nature of the activity. Ask the provider what records it can supply before payment milestones make your bargaining position weaker.
Use five control points through the project cycle
First, the technical lead defines the project scope and separates experimental work from routine production, maintenance or ordinary commercial services. The tax reviewer then assesses the relevant eligibility questions using the actual activity, not a marketing label.
Second, finance creates project-level cost tracking with categories and source documents. Employees know which time, procurement and service records they need to maintain. Data should be proportionate and contemporaneous, rather than reconstructed solely at filing time.
Third, the contract owner checks outsourced arrangements and the identity of the claiming taxpayer. Related-party involvement, overseas work or shared intellectual property should be flagged for specific advice. Do not assume that every group company can claim the same expense.
Fourth, accounting and tax staff reconcile the project schedule to the general ledger and the treatment of any resulting intangible asset. Differences are explained and approved. A technical milestone and an accounting capitalisation decision are related but not interchangeable.
Fifth, management reviews a cash-flow forecast that shows operating expenditure, committed financing and potential tax effects separately. It tests a scenario in which a disputed cost is excluded or the expected benefit is not immediately usable. The project decision should remain understandable under that scenario.
Keep the claim narrower than the evidence
If the technical and financial records support only part of the expenditure, the response should be a targeted review of that part. Do not inflate documentation with invented experiment logs or relabel ordinary costs to fit a desired saving. Where uncertainty is material, obtain a written professional assessment before relying on the benefit.
This guide does not determine eligibility or calculate your company’s tax saving. Industry exclusions, expense categories, outsourced arrangements, accounting treatment and the taxpayer’s position can alter the result. National tax relief also should not be confused with local grants carrying different application and repayment conditions.
The practical outcome is an expense trail that another reviewer can follow: activity, cost, entity, treatment and evidence. That trail helps both tax compliance and project management. It allows your business to compare the real cost of development options without treating a tax incentive as guaranteed financing for an otherwise unfunded project.
