Free Trademark Search in Nepal: How to Search the DOI Register
A free trademark search in Nepal is possible: this guide covers the DOI Bulletin records, the layered search workflow, and what a free search can and cannot see.
Before any fee is paid or any signboard printed, one question decides whether a brand name is worth filing in Nepal: is it already taken? The good news is that a useful trademark search in Nepal costs nothing but care — the register's public records are searchable if you know where they live.
This guide walks the free trademark search Nepal businesses can run themselves: the sources, the layered workflow, and the honest limits of what a free search can see.
In short
A free trademark search in Nepal starts with the Department of Industry's Industrial Property Bulletin — PDF issues dating back to 2006 that list each application with mark, applicant, class, goods and dates in BS and AD. Structured datasets of 71,000+ published records make the same search instant. The workflow: search the exact name, then phonetic and spelling variants, then Devanagari transliterations, then filter by class and read the goods wording. A free search cannot see pending unpublished applications — treat it as a screen, not a guarantee.
What a free trademark search in Nepal actually searches
The DOI's Industrial Property Section publishes the Industrial Property Bulletin, and each issue lists trademark applications accepted for publication — the mark, the applicant, the class, the goods or services, filing and publication dates, and the application and record numbers. Issues date back to 2006 and print dates in Bikram Sambat as well as AD.
What the Bulletin is not: a live, indexed register with instant fuzzy search. Records arrive as PDFs, extraction introduces letter-spacing artefacts, and pending applications that have not yet published are invisible. That gap between 'published record' and 'full register' is the single most important thing to understand about searching in Nepal.
The free search workflow, layer by layer
A competent free search proceeds in layers, because each layer catches conflicts the previous one misses.
Search the exact name
Run the mark as spelled across both name fields the records carry. Exact hits are the high-certainty conflicts — and the easy ones to find.
Search phonetic and spelling variants
Section 18(1) turns on deceptive similarity, not identity: PH/F, C/K, doubled letters, plurals and Y/I endings all generate live conflicts. Search the variants you would be annoyed to be opposed by.
Search transliterations
In a bilingual market the same brand can appear as EVEREST, एभरेष्ट or a romanisation of it. Cross-script variants are where exact-match searching fails completely — and where most DIY searches stop too early.
Filter by class, then read the goods
Class narrows the field; the goods wording decides it. A same-class hit for overlapping goods is a conflict; a same-class hit for unrelated goods may not be. Read the specification, not just the class number.
Read each hit as a record
For surviving candidates, capture the mark and applicant, class and goods, application number, filing and publication dates in BS and AD, and the Bulletin issue — the publication date is what starts the 90-day opposition clock.
Reading a record: the fields that matter
Each published record rewards a specific reading, and two fields do most of the work.
| Field | Why it matters |
|---|---|
| Mark name / display text | The conflict surface; device marks may differ from their text field |
| Applicant | Maps a competitor's whole portfolio when searched across issues |
| Class + goods | The claimed scope of protection under Section 18A |
| Filing date | First-to-file priority arguments |
| Publication date (BS + AD) | Starts the ~90-day opposition window |
| Bulletin + record number | Traceable citation for any opposition or clearance note |
What a free search cannot see
Three blind spots are structural. Unpublished pendings — applications filed but not yet published appear nowhere; there is always a window of invisibility. Extraction artefacts — PDF-sourced text can carry letter-spacing damage that defeats naive matching. Interpretation — a search returns records; whether two marks are 'deceptively similar' for the same goods is a judgement Section 18(1) leaves to examination and argument.
The practical response is calibration: treat a clear multi-layer search as strong but not absolute comfort, and treat any same-class near-hit as a question worth a professional opinion before spending on filing and launch.
When free is enough — and when to pay
Free is enough when it answers the cheap question: is this name obviously taken, obviously free, or worth a lawyer's hour? Most names are one of the first two, and the structured tools answer them in seconds. Pay when the answer is arguable — a same-class phonetic near-miss, a Devanagari/Latin pair that could transliterate into each other, or a launch plan with real marketing money behind it.
An owner search adds a second free layer: searching a competitor's name across all issues maps their filing pattern — every mark, every class, every date — which is strategy, not trivia.
Want this handled for your brand?
IP Watch monitors new trademark publications in Nepal and alerts you to potentially conflicting marks — with the context needed to review them.
This article is general information, not legal advice.