China Trademark Similarity Search: Avoid Costly Filing Mistakes
| Key Fact | Figure | Why It Matters |
|---|---|---|
| Effective trademark registrations in China | Over 51 million | The registry is more crowded than ever, raising conflict risk for every new filing |
| Rejections traced to identical or similar prior marks | Over 87% | Similarity conflicts are the single biggest cause of trademark rejection in China |
| Rejection review overall success rate (2025 data) | Under 19% | Fixing a rejection after the fact is far harder than avoiding it through search |
Filing a trademark in China without a proper similarity search is one of the most common — and most expensive — mistakes foreign businesses make. With China’s trademark registry now holding more than 51 million active marks, the odds that your proposed brand name, logo, or slogan overlaps with something already on file are higher than most applicants expect.
This guide explains how China National Intellectual Property Administration (CNIPA) examiners actually evaluate similarity, what the underlying law requires, and how to search effectively before you file — so you understand the real conflict risk before committing time and money to an application. For a broader walkthrough of the search process itself, see YCIP’s complete guide to China trademark search.
Why Similarity Search Matters More Than Ever in 2026
A Record-Crowded Trademark Registry
As of mid-2026, China’s effective trademark registrations have surpassed 51.97 million marks [1]. In the second quarter of 2026 alone, CNIPA received roughly 3.4 million new trademark applications while granting about 2.0 million registrations [1]. The national average rejection rate has climbed year over year, and industry estimates now place it in the roughly one-third to 35% range, with saturated classes such as clothing, food and beverage services, and advertising running notably higher [1][6].
This scale matters directly for similarity search: the more marks that already exist, the higher the statistical chance that any new name, logo, or slogan collides with something already registered. Industry data indicates that more than 87% of all rejections stem from conflicts with identical or similar prior marks [1]. For short marks — particularly two- and three-character combinations — the rejection rate can approach 90% [1].
The Real Cost of Skipping a Proper Search
Applicants who file without adequate clearance searching are not just risking rejection — they are risking a specific, quantifiable set of losses:
| Cost Category | Estimated Impact |
|---|---|
| Application fees wasted | RMB 270 official fee per class, plus RMB 700–1,200 agency service fees per class |
| Time lost | Roughly a 4-month examination cycle per rejected application [6] |
| Branding disruption | Delayed market entry and potential forced rebranding costs |
| Opportunity cost | Competitors may secure similar marks while you restart the process |
The upside is equally clear: applicants who complete a thorough similarity search, verify classification, and prepare documentation carefully can reduce their rejection risk well below the national average, while applications handled by non-professional agents tend to show meaningfully higher rejection rates [1]. For a full breakdown of official and agency costs, see YCIP’s China trademark registration fee guide.
How China Defines Trademark Similarity: The Legal Framework
The Core Standard: Article 30 of the Trademark Law
The primary legal basis for similarity-based rejection is Article 30 of the Trademark Law of the People’s Republic of China (《中华人民共和国商标法》第三十条). In substance, it provides:
“Any trademark application identical with or similar to a trademark already registered or preliminarily approved by another party, for use on the same or similar goods, shall be rejected and not published.” [2]
This provision applies whether the prior mark is already registered or merely preliminarily approved — meaning a pending mark can block your application even before it completes registration. Chinese courts and CNIPA examiners interpret the article according to several consistent principles [2][3]:
- Similarity is determined by whether confusion among relevant consumers is likely, not by a technical or side-by-side comparison
- Examiners reference the official Classification of Similar Goods and Services (《类似商品和服务区分表》) while also weighing actual use in the marketplace
- Well-known trademarks may receive cross-class protection even where the goods or services are not identical
Related Provisions Applicants Should Know
Article 30 rarely operates in isolation. Several related provisions shape the broader similarity landscape:
| Provision | Relevance to Similarity |
|---|---|
| Article 31 | Governs conflicts between same-day applications; frequently cited alongside Article 30 in rejection reviews |
| Article 10 | Absolute grounds for refusal (national symbols, deceptive marks); not similarity-based but relevant to full pre-filing screening |
| Article 11 | Lack of distinctiveness (descriptive or generic terms); often confused with similarity issues but legally distinct |
| Article 57 | Defines infringement for unauthorized use of similar marks on similar goods post-registration |
Applicants should also account for the revised Trademark Law, adopted by the Standing Committee of the National People’s Congress on June 26, 2026, which expands the statute from 73 to 87 articles across nine chapters and takes effect January 1, 2027 [5]. While the revision does not change the substantive similarity test under Article 30, several related changes affect the broader filing environment: an explicit bar on applications “not intended for use and clearly exceeding normal business needs,” a shortened opposition period cut from three months to two, and stronger accountability for filing agencies found to knowingly assist bad-faith applications [5]. Because the current law still governs applications filed before 2027, YCIP frames all similarity guidance under the applicable version at the time of filing — see YCIP’s coverage of China’s non-use cancellation rules for a related compliance mechanism under the current framework.
What CNIPA Examiners Actually Look For
The “Sound, Shape, Meaning” Test for Word Marks
For word marks, CNIPA examiners apply a three-dimensional comparison defined in the Trademark Examination and Adjudication Guidelines (《商标审查审理指南》) [3]:
| Dimension | Assessment Criteria |
|---|---|
| Shape (形) | Similar character composition, fonts, or arrangement |
| Sound (音) | Identical or similar pronunciation |
| Meaning (意) | Same or similar semantic meaning |
Critically, a match on any single dimension can trigger rejection — the mark does not need to fail all three. For marks of three characters or more, examiners have consistently found similarity even when only individual characters differ and there is no meaningful change in overall meaning [3].
Visual Impact for Figurative and Combined Marks
Figurative (图形) and combined (图文组合) marks are evaluated differently but no less strictly. Examiners weigh three factors [3]:
- Overall visual impact — composition, style, and color scheme. Minor stylistic adjustments, such as a slightly altered curve in a logo, do not overcome a similar overall impression.
- Core identifying elements — if a mark’s central recognizable element overlaps with a prior registration, decorative additions will not save the application.
- Dual scrutiny for combined marks — for word-plus-graphic combinations, examiners assess the word and graphic elements separately. A conflict in either element can sink the entire application, regardless of how distinct the other element is.
This dual-scrutiny rule explains a counterintuitive but well-documented pattern: an analysis of 7,059 successful rejection-review cases in 2025 found that the two leading reasons for overturning a rejection were a finding of no similarity (42.90% of successful cases) and the lapse of the cited prior mark (33.00%) — underscoring how central the similarity determination is to the entire examination process [4].
The “Ordinary Consumer” Perspective
Examiners do not evaluate similarity the way a trademark attorney would, through careful side-by-side analysis. Instead, they apply the standard of the “ordinary consumer’s general attention” (相关公众的一般注意力) [3]:
- Examiners simulate a quick, real-world recognition scenario — not a detailed professional comparison
- Comparison is conducted in isolated settings, meaning marks are assessed as a shopper would encounter them separately, not side by side
- The distinctiveness and market reputation of the prior mark are also factored in
The practical implication is significant: even where an applicant can clearly distinguish two marks upon close inspection, the mark will still be rejected if an ordinary consumer glancing at a shelf would likely be confused about the source. This is why professional judgment — not just a database lookup — is central to a reliable clearance opinion. See YCIP’s guide on proving trademark distinctiveness in China for related strategy on strengthening a mark’s position.
How to Conduct an Effective Similarity Search
The Search Blind Period: Why Even a Perfect Search Can Miss Conflicts
One of the most underappreciated risks in China trademark clearance is the “search blind period” (查询盲期). After an application is submitted, it typically takes one to three months before it appears in any searchable database — official or professional [6]. During that window, newly filed applications are effectively invisible to everyone, including CNIPA examiners performing prior-mark checks on later filings.
With daily new trademark filings in China approaching 20,000 [1], the blind period represents a real and unavoidable exposure: hundreds of thousands of pending applications may sit outside any searchable index at a given moment. No amount of search diligence eliminates this risk entirely — it can only be managed through early filing and periodic re-screening. YCIP’s guide on using the CNIPA trademark search system walks through the practical mechanics of running these checks yourself.
Official Tools vs. Professional Search Platforms
Applicants generally choose among three tiers of search capability:
| Search Type | Coverage | Best For |
|---|---|---|
| CNIPA official website (中国商标网) | Basic, free, official data | Initial self-screening |
| Professional databases | Broader coverage, sound-shape-meaning matching, AI-assisted analysis | Comprehensive professional searches |
| Expert human review | Combines database results with examination-standard knowledge and industry experience | High-value brands, complex combined marks |
CNIPA’s own official search platform has continued to expand its capability, including an AI-powered “search by image” feature launched in 2026 [6]. These tools are genuinely useful for a first pass, but they were never designed to replace the judgment of someone who understands how examiners actually apply the sound-shape-meaning test. For a comparison of specific platforms, see YCIP’s guide to the best trademark search engines in China and database search guide.
The Search Depth That Actually Works
Search thoroughness correlates directly with approval outcomes. Industry data illustrates a clear pattern:
| Search Depth | Estimated Success Rate |
|---|---|
| Basic preliminary screening only | ~5% |
| Full database search with sound-shape-meaning comparison | ~70–80% |
| Expert human assessment against examination standards and industry practice | ~85–90% |
A complete similarity search should cover identical marks in the same class, similar marks in the same class across sound, shape, and meaning, identical or similar marks in related classes, and any well-known marks that may carry cross-class protection. YCIP’s CNIPA search mastery guide and brand name availability guide go deeper into structuring a search of this scope.
What Happens If a Similarity Issue Is Found
Your Options When Facing a Prior Similar Mark
Discovering a potential conflict before filing is far better than discovering it after rejection, because it leaves more options on the table:
- Abandon and redesign — often the most cost-effective path when the prior mark is strong and directly relevant
- Narrow the goods or services — limit the application to sub-categories where the conflict may not apply
- Argue non-similarity — build a case on sound-shape-meaning differences, supported by use evidence
- Acquire the prior mark — purchase the existing registration through a trademark assignment
- Wait for the prior mark to lapse — if it appears vulnerable to non-use, pursue a non-use cancellation (撤三) action
Each path carries different cost, timeline, and risk trade-offs, and the right choice depends heavily on how central the trademark is to your brand strategy. See YCIP’s guide on trademark squatting in China for a related scenario where acquisition or cancellation strategies often come into play.
Rejection Review: Success Rates and Realities
For applicants who do receive a similarity-based rejection, rejection review (驳回复审) is available — but the odds are sobering. Based on 2025 data covering 190,655 cases reviewed under Articles 30 and 31 [4]:
| Outcome | Number | Percentage |
|---|---|---|
| Full preliminary approval | 7,059 | 3.70% |
| Partial preliminary approval | 29,105 | 15.26% |
| Total success rate | 36,164 | 18.97% |
An overall success rate under 19%, with full approval achieved in less than 4% of cases, makes one point unmistakable: getting the search right before filing matters far more than hoping to win on review afterward. For applicants already facing a rejection, YCIP’s guides on appealing a similarity-based refusal and understanding a CNIPA rejection notice outline the review process in detail.
Practical Tips to Improve Your Approval Odds
Bringing the data above together, a few practices consistently separate applicants who clear examination from those who don’t:
- Search before you commit to branding — run a similarity search before investing in packaging, marketing, or domain names built around the proposed mark
- Search across all three dimensions — sound, shape, and meaning — not just an exact-text match
- Check related classes, not just your primary class — cross-class conflicts are a common blind spot
- Account for the search blind period — consider a second screening shortly before filing, and again after your application clears initial examination
- Evaluate combined marks separately — a compliant graphic paired with a conflicting word element (or vice versa) will still be rejected
- Get a professional risk opinion before filing — especially for marks central to long-term brand identity, where a rejection is costly to absorb
Given that non-use cancellation, assignment negotiations, and rejection review all carry meaningfully lower success rates than a well-executed pre-filing search, the search stage is where applicants get the best return on legal spend. YCIP regularly supports foreign applicants through this process — see why hiring a China trademark lawyer matters for a broader view of where professional support adds value.
Frequently Asked Questions
What is the difference between “identical” and “similar” trademarks under Chinese law?
Under the Trademark Examination and Adjudication Guidelines, identical means two marks are completely the same or differ only in minor ways — such as font, letter case, spacing, or color [3]. Similar means the marks are not identical but are close enough in character shape, pronunciation, meaning, or overall structure that they could cause consumer confusion. Identical marks face automatic rejection; similar marks are evaluated case by case for confusion likelihood.
How much does a China trademark similarity search cost?
A CNIPA official website search is free but limited in scope. Basic professional searches are often bundled into agency filing service fees, typically around RMB 700–1,200 per class, on top of the RMB 270 official application fee per class. Comprehensive searches with expert sound-shape-meaning analysis and conflict risk assessment are priced separately and vary by firm. Be cautious of unusually cheap “all-inclusive” offers — these frequently cannot even cover the official government fee and often result in superficial searches.
Can I register a similar trademark if I add a distinctive element?
Generally, no. Examiners apply what practitioners describe as a “prominent part association” rule. For combined marks, if either the word or graphic element matches a prior registration, the entire application can be rejected regardless of how distinctive the added element is. Adding descriptive or generic terms to a word mark typically does not overcome a similarity finding either — the test centers on whether the core identifying elements could confuse consumers.
What is the “search blind period” and how does it affect my application?
The search blind period is the one-to-three-month window between when a trademark application is filed and when it becomes visible in any searchable database. During this period, no one — including CNIPA examiners — can search for these pending applications. With filing volumes approaching 20,000 per day, this creates an unavoidable pocket of risk that no search methodology can fully eliminate.
What is the success rate of rejection review for similar trademarks?
Based on 2025 data covering 190,655 cases rejected under Articles 30 and 31 that entered review, the overall success rate was 18.97%, with full preliminary approval achieved in only 3.70% of cases [4]. The most common reasons for a successful review were a finding of no similarity and the lapse of the cited prior mark — reinforcing that pre-filing search is far more effective than relying on post-rejection appeals.
Conclusion: Search First, File with Confidence
China’s trademark registry has never been more crowded, and the data is unambiguous: similarity conflicts drive the overwhelming majority of rejections, while post-rejection remedies succeed less than one time in five. A thorough, professionally guided similarity search — one that checks sound, shape, and meaning across related classes, accounts for the search blind period, and applies the same standards CNIPA examiners use — is the single highest-leverage step you can take before filing in China.
YCIP has guided foreign businesses through China trademark clearance, filing, and appeal for years, combining database search with the examination-standard judgment that database tools alone cannot replicate. If you’re preparing to file — or want a second opinion on a mark you’re already committed to — explore YCIP’s trademark and copyright services, or contact our team directly to request a similarity search and filing quote.
Disclaimer
This article is provided for general informational purposes only and does not constitute legal advice. Trademark examination outcomes depend on the specific facts of each case. For advice on a specific application or conflict, consult a qualified China trademark attorney.
References
- “CNIPA Quarterly Statistics and Trademark Registration Data 2026”, https://english.cnipa.gov.cn/. Source Role: Official government statistics. Support Status: supports. Relevance: Basis for registration totals, quarterly filing volume, and rejection rate figures.
- “Trademark Law of the People’s Republic of China, Article 30”, https://english.cnipa.gov.cn/. Source Role: Primary statute. Support Status: supports. Relevance: Core legal standard for similarity-based rejection.
- “Trademark Examination and Adjudication Guidelines (商标审查审理指南)”, https://english.cnipa.gov.cn/. Source Role: Official examination guidance. Support Status: supports. Relevance: Basis for the sound-shape-meaning test and ordinary consumer standard.
- “2025 Rejection Review Case Data Analysis”, industry legal analysis. Source Role: Industry/legal commentary. Support Status: supports. Relevance: Basis for rejection review success rate statistics.
- “China’s 2026 Trademark Law Revision: A Shift Toward a Use-Oriented Trademark Regime”, https://www.loeb.com/en/insights/publications/2026/07/chinas-2026-trademark-law-revision-a-shift-toward-a-use-oriented-trademark-regime. Source Role: Legal analysis. Support Status: supports. Relevance: Confirms adoption date, article count, and effective date of the revised Trademark Law.
- “How to Use CNIPA Trademark Search Online Like a Professional”, https://yciplaw.com/how-to-use-cnipa-trademark-search-online-like-a-professional/. Source Role: Internal reference. Support Status: supports. Relevance: Basis for search blind period, examination cycle, and search tool details.



