BUY WEBSITES
BUY DOMAINS
LOGIN
SIGNUP
USPTO Trademark Search Walkthrough: How to Clear a Domain Name Against the Federal Register Before You Buy It
The USPTO trademark search is the free, official way to check whether a brand name is already registered as a federal trademark in the United States before money changes hands. The United States Patent and Trademark Office runs the register, and it lets anyone query it through the Trademark Search system at no cost.
For a domain buyer the search answers one practical question. Does the name on this aged or expired domain collide with a live federal trademark in the category the buyer intends to use it for? A clean result lowers conflict risk. A live exact match in the same class is a red flag worth pricing in before purchase.
This walkthrough runs the search the way a careful buyer runs it, tool by tool and field by field, then sets out how to read the result and what an official search will not catch. SEO Domains operates the curated marketplace where an aged or expired domain’s name is vetted before it is listed, so the brand string is screened before it is bought, not after.
What the USPTO trademark search is, and why a domain buyer runs it
The USPTO trademark search is a free query against the United States federal trademark register, run through the Trademark Search system at tmsearch.uspto.gov. A domain buyer runs it to confirm that the brand string on an aged or expired domain does not collide with a live federal trademark in the goods or services category the domain will be used for.
The United States Patent and Trademark Office, the USPTO, is the federal agency that registers trademarks. It maintains the public register and exposes it through search tools that cost nothing to use. The headline question for an SEO or domain investor is not academic. It is whether the name attached to a domain carries a trademark-conflict liability that a trademark owner can act on.
The domain-buyer reading of the search
A brand owner runs the USPTO search to clear a name before filing an application. A domain buyer runs the same search for a related reason. An aged or expired domain frequently carries a real brand string from a prior business, and that string can match a trademark someone else still holds. Buying the domain does not transfer that trademark, and using the name commercially in a colliding class invites a dispute.
The dispute mechanisms are documented elsewhere in this hub: the federal courts under the Lanham Act, and the domain-specific routes covered in the cybersquatting law hub and the UDRP proceedings hub. The USPTO search is the cheapest first filter that keeps a buyer out of those mechanisms in the first place.
The aged-domain angle most searchers miss
An expired domain is a special case the brand-owner guides never address. The name on a dropped domain frequently belonged to a company that filed, or never filed, a trademark, and the two facts move independently. A business can abandon a domain while keeping its live trademark, so the name reads as available on a registrar yet remains locked in the register. The reverse also happens: a name nobody held when the domain first registered can carry a trademark filed by a newer company years later. The search reads the register as it stands today, which is the only state that governs a purchase made today.
The three USPTO tools you will use, plus the international layer
The federal search runs across three USPTO tools. The Trademark Search system is the register itself. TSDR pulls the full status and document file for a specific mark. The Trademark ID Manual defines the goods and services identifications that map a domain to its class. Beyond the federal register, the WIPO Global Brand Database adds the international layer.
Trademark Search: the official federal register
The Trademark Search system at tmsearch.uspto.gov is the USPTO’s own database of registered and applied-for trademarks. The USPTO retired the Trademark Electronic Search System, TESS, on 30 November 2023 and replaced it with this cloud-based system, which the office built to be faster and more stable after TESS had run since February 2000. A query runs as a basic search against the mark text, or as an advanced search that adds field tags, wildcards, boolean operators, and status filters.
TSDR: the full file on one mark
The Trademark Search system shows the register hit. TSDR, Trademark Status and Document Retrieval, opens the full file behind that hit. Enter a serial number or registration number and TSDR returns the live status, the prosecution history, the owner, the goods and services, and the documents filed in the case. A domain buyer uses TSDR to confirm whether a worrying match is a live registration, a pending application, or a dead mark, which is the single fact that decides how much the hit matters.
The ID Manual: scoping the class before you search
Trademarks are registered for a name within categories of goods and services, defined by the Nice classification, an international system of 45 classes. The Trademark ID Manual is the USPTO tool that lists acceptable descriptions of goods and services and the class each falls in. A buyer uses it to map the domain’s intended use to its correct class before searching, because the class is the difference between a conflict and a non-conflict.
The international layer: WIPO Global Brand Database
The federal register is the US picture. A name can be clear in the United States yet held abroad. The WIPO Global Brand Database, run by the World Intellectual Property Organization, aggregates trademark records from national, regional, and international collections, including marks filed under the Madrid System. The WIPO database lets a buyer extend the same brand-string check beyond the USPTO when the domain will trade internationally, and it supports text and image search across those collections.
| Tool | What it searches | Why a domain buyer uses it |
|---|---|---|
| Trademark Search (tmsearch.uspto.gov) | The official US federal trademark register | The core check for the exact and similar name across the federal register |
| TSDR | The full status and document file for one serial or registration number | Confirms whether a hit is live, pending, or dead, and reads the owner and class |
| Trademark ID Manual | Acceptable goods and services and their Nice classes | Pins the domain’s intended use to the correct class before searching |
| WIPO Global Brand Database | Records from national, regional, and international collections | Extends the check abroad for a domain that will trade internationally |
Before you search: pin the brand string and the trademark class
A useful USPTO search starts before the first query. The buyer fixes two inputs: the exact brand string to test, including the variants a register treats as the same mark, and the class that matches the domain’s intended commercial use. Searching without these two inputs produces noise instead of an answer.
Pin the string and its variants
The brand string is the second-level part of the domain name, read as a word and not as a URL. A register treats a word mark loosely, so the search has to cover the variants that count as the same mark: spacing, hyphenation, common misspellings, and obvious phonetic equivalents. A domain reading brightpath is tested as brightpath, bright path, and bright-path, because a register would read those as one mark. The wildcard operators in the next section are what let one query cover that spread.
Pin the trademark class
The class scoping is what turns a flood of irrelevant hits into a focused answer. The buyer decides what the domain will sell or serve, maps that to its class through the ID Manual, and searches with that class in mind. A domain destined for an affiliate site about running shoes lives in class 25 and class 35. A software tool lives in class 9 and class 42. The class is the filter that separates a real conflict from an unrelated mark that shares the same word.
Input 1: the brand string and its variants
The domain word read as a brand, plus its spacing, hyphenation, misspelling, and phonetic forms. A register reads brightpath, bright path, and bright-path as one mark, so one query has to cover them all.
Input 2: the intended trademark class
The Nice class or classes that match what the domain will sell or serve, mapped through the ID Manual. The class is what makes a same-word hit either a real conflict or an unrelated mark.
The step-by-step USPTO trademark search walkthrough
The federal search runs in five steps: a knockout basic search on the exact string, an advanced search that adds field tags and the class, a wildcard pass for the variants, a status filter to drop dead marks, and a read of the full file in TSDR. Each step narrows a broad name down to the handful of records that govern the purchase. This is the sequence a careful buyer follows every time.
The walkthrough below uses the post-TESS Trademark Search system. The operator examples, field tags, and the wildcard syntax are drawn from the USPTO’s own search documentation and the Alt Legal guide to the new system. A junior buyer can follow it field by field.
-
Run a knockout basic search on the exact string
Open the Trademark Search system at tmsearch.uspto.gov and enter the brand string in the basic search. This is the knockout pass: it surfaces the obvious, identical-name hits fast. Read the count first. A common word returns thousands of marks, which is the signal to escalate to the advanced search and the class filter instead of scrolling.
The mistake: stopping at the basic search. A clean basic result feels reassuring, but it misses spelling variants and same-class similar marks. The knockout pass is the start of the search, never the whole of it.
-
Switch to advanced search and add the class field tag
Move to the advanced search and add the class with the international-class field tag, written as three digits: IC025 for class 25, not IC25. Combine it with the brand string using the AND operator, so a query reads as the string AND IC025. The Alt Legal walkthrough documents narrowing a 13,000-hit search down to 21 records by stacking the class and live-status filters this way.
The mistake: searching the name with no class. Every same-word mark in every one of the 45 classes lands in the result, and the one mark that matters to the buyer’s category is buried under registrations that cannot collide with it.
-
Add wildcards to catch the variants
Use the truncation wildcard to cover the spelling spread in one query. A period-asterisk after a stem, written as cat.*, matches cat, cats, and catalog; placed before a stem it widens the other way. Regular-expression searches are wrapped in forward slashes. One wildcard query covers the variants that three separate exact searches would otherwise need.
The mistake: running only the exact string. A register reads brightpath and bright-path as one mark, and an exact-only search reports a domain as clear when a hyphenated registration of the same word sits live in the same class.
-
Filter to live marks with the status tag
Drop the dead records by adding the live-designation tag, written as LD:true, to the query. A dead or abandoned mark is a far lower risk than a live one, so filtering to live first focuses the read on the records that can still be enforced. Keep one pass without the filter, though, because a recently dead mark can signal a name a prior owner cared enough to register.
The mistake: treating every hit as equal. A search that does not separate live from dead reads a long-abandoned registration as the same threat as a live one, and the buyer either panics at a dead mark or overlooks the live one beneath it.
-
Open the full file in TSDR and read the owner and class
Take the serial or registration number of any live, same-class hit into TSDR. The full file confirms the status, the owner, the exact goods and services, and the prosecution history. This is the read that turns a register match into a decision: a live registration on the same word in the same class, held by an active owner, is the finding that prices a domain down or out. The diligence on the domain’s own history runs in parallel through the expired domain fundamentals hub.
The mistake: deciding from the result list alone. The list shows the mark; only TSDR shows whether it is live, who owns it, and exactly what it covers. A buyer who skips the file read acts on half the facts.
How to read a search result: live versus dead, applied versus registered
A USPTO result is read on four fields: the live-or-dead status, whether the mark is applied for or registered, the class, and the owner. A live registration on the same word in the same class is the high-risk finding. A dead mark, an unrelated class, or a pending application that has not yet registered each lowers the weight of the hit.
Live versus dead
Status is the first field to read. A live mark is in force and can be enforced by its owner. A dead mark has been abandoned, cancelled, or expired, and it carries far less weight, though a recently dead mark on the exact domain string is worth noting as a sign the name once mattered to a business. The Trademark Search system marks status on the record, and TSDR confirms it on the full file.
Applied for versus registered
The second field is the stage. A registered mark has cleared examination and publication and holds the full set of rights. A pending application is a claim in progress that has not yet registered, and it faces refusal, opposition, or abandonment before it does. A pending application on the buyer’s string in the buyer’s class is a real signal, but a softer one than a registration, because its outcome is not yet settled.
Class and owner
The third and fourth fields decide whether the live, registered hit truly reaches the buyer. The class confirms whether the mark covers the domain’s intended category or an unrelated one. The owner shows who holds the right and whether they are an active business that would defend it. A live registration in an unrelated class, or one held by a long-dissolved entity, is a weaker threat than a live registration in the same class held by an operating company.
| Field | What it tells the buyer | Higher risk reading |
|---|---|---|
| Status (live or dead) | Whether the mark can still be enforced | Live and in force |
| Stage (registered or pending) | Whether the right is settled or still in progress | Registered and past publication |
| Class | Whether the mark covers the domain’s category | Same or coordinated class |
| Owner | Who holds the right and whether they would defend it | An active, operating business |
Knockout versus comprehensive searching: the escalation
A federal search escalates in two stages. The knockout search catches identical and near-identical marks fast and rules out the obvious collision. The comprehensive search reaches further, into phonetic equivalents, alternate spellings, translations, and design elements, where a confusingly similar mark hides. A domain buyer runs the knockout first, then escalates when the name carries real value.
The knockout search
The knockout search is the first pass. It tests the exact string and its closest variants against the register to rule out a direct, identical-name collision. For a low-value domain or a name with no commercial brand reading, a clean knockout is a proportionate level of diligence. It is fast, free, and catches the collisions that carry the highest weight.
The comprehensive search
The comprehensive search is the escalation. Trademark conflict is governed by likelihood of confusion, not by spelling, so a name can collide with a mark it does not literally match. The comprehensive pass tests phonetic equivalents, alternate and foreign-language spellings, translations, and the design-code layer for figurative marks. It is the level professional clearance searches reach, and the level worth running before building a brand on a higher-value aged domain.
Matching the search depth to the domain
The right level of search scales with what the domain is worth and how it will be used. A cheap utility domain destined for a non-branded affiliate site warrants a knockout pass. A premium aged domain that will carry a real brand warrants the comprehensive search, and at the top end, a clearance opinion from trademark counsel. The search is a graded filter, and matching its depth to the stakes is the discipline that separates real diligence from a box ticked.
What the USPTO search will not catch
A clean federal search lowers risk; it does not grant a right or guarantee safety. The USPTO register does not cover unregistered common-law marks, state trademark registrations, or foreign marks. It is not a legal opinion on likelihood of confusion. These are the limits to recognise, so a clean result is read as the filter it is, not as a clearance it is not.
Common-law and state marks
Trademark rights in the United States arise from use, not only from federal registration. A business that has used a name in commerce without ever filing holds common-law rights in its market, and those rights never appear in the federal register. State trademark registrations sit in separate state databases the USPTO search does not reach. A name that is clean federally can still collide with a common-law user or a state registration the federal search never showed.
The international gap and the figurative gap
The federal register is national. A name clear in the United States can be held abroad, which is the gap the WIPO Global Brand Database fills for a domain that will trade internationally. The register also indexes word marks more readily than figurative ones. A logo-based mark with a stylised word can sit in the design-code layer a text-only search skims past, which is why the comprehensive search reaches into design codes.
The search is a filter, not a verdict
The deepest limit is the nature of the tool. The USPTO search reports what is on the register. It does not rule on likelihood of confusion, the legal test that decides an actual conflict, and only a court or trademark counsel does that. The search documents diligence and prices risk; it does not settle the legal question. A buyer who reads a clean result as a green light has mistaken a filter for a verdict.
The domain due-diligence checklist: red flags before you buy
The federal search resolves to a short decision list. Seven findings cover the readings a domain buyer meets, and each maps to a clear action: buy with confidence, watch and price the risk in, or walk away. The high-risk rows share one shape every time, a live registered mark on the same word in the same class, held by an active owner. Use this as the scannable reference.
The table below consolidates the readings scattered through the walkthrough and the limits sections into one decision layer. The left column is the search finding, the centre column is what it means for the purchase, and the right column is the action it calls for. Read top to bottom, the rows run from the cleanest result to the one that ends a deal.
| Search finding | What it means for the purchase | Action |
|---|---|---|
| No hit on the string in any class | The federal register is clear of the name; the lowest-friction result | Buy: clear federally, extend to common-law for a high-value name |
| Dead or abandoned mark on the string | A prior registration that is no longer enforceable; lower risk | Buy with note: confirm the dead status in TSDR before relying on it |
| Live mark in an unrelated class | The same word owned for a category the domain will not enter | Buy if classes stay clear of each other; keep the use in its lane |
| Pending application in the same class | A claim in progress that has not registered and may not | Watch: track in TSDR; the risk firms up only if it registers |
| Live registered mark in a coordinated class | A neighbouring category that can still reach the domain’s use | Watch: test the overlap; price the conflict risk into the offer |
| Live registered mark, same word, same class | A direct collision in the domain’s intended category | Walk: the name carries an enforceable conflict for this use |
| Live mark held by a large active brand | An owner with the resources and motive to enforce | Walk: the enforcement risk is highest where the owner is active |
One pattern runs down the action column. The findings that walk a buyer away from a domain all share a shape: a live registration, on the same word, in the same class, held by an active owner. The findings that clear a domain share the opposite shape: dead status, an unrelated class, or no hit at all. A buyer who reads those four fields on every hit has done the diligence the USPTO search exists to provide, and sourcing from a marketplace that screens the name first means starting the check from a shorter, cleaner list.
USPTO trademark search frequently asked questions
The five questions domain buyers and SEOs raise when they run a USPTO trademark search, answered against the USPTO’s own tool documentation and the use-based nature of trademark rights in the United States.
Q1Is the USPTO trademark search free, and can I run it myself?
Yes on both. The Trademark Search system at tmsearch.uspto.gov is a free public tool, and TSDR and the ID Manual are free too. Anyone can run a knockout search themselves without an account. The cost layer enters only when a buyer files an application or engages counsel for a comprehensive clearance opinion on a high-value name.
Q2What happened to TESS, and is my old search guide still valid?
The USPTO retired TESS, the Trademark Electronic Search System, on 30 November 2023 and replaced it with the cloud-based Trademark Search system. TESS had run since February 2000. Any guide that still references TESS screens or shortcuts is out of date: the new system uses field tags such as the three-digit class tag and the LD:true live filter, plus wildcard and boolean operators that the old system handled differently.
Q3Does a clean USPTO search mean a domain name is safe to use?
No, and this is the limit to understand. A clean federal result rules out registered and applied-for federal marks, but trademark rights in the United States arise from use, so a common-law user or a state registration the federal search never shows can still hold a claim. The search lowers risk and documents diligence; it does not grant a right or settle likelihood of confusion, which only a court or trademark counsel decides.
Q4How does the trademark class change the result of a search?
A trademark protects a name only within its registered classes, of which there are 45 under the Nice system. A name registered for clothing in class 25 does not block the same name used for software in class 42. Scoping the search to the domain’s intended class, mapped through the ID Manual, separates a real conflict from an unrelated mark that shares the word. Coordinated classes then test whether a neighbouring category reaches into the buyer’s use.
Q5Why run a USPTO search before buying an aged or expired domain?
An aged or expired domain frequently carries a real brand string from a prior business, and that string can match a trademark someone still holds, or one a newer company filed after the domain first registered. Buying the domain does not transfer the trademark, so using the name commercially in a colliding class invites a dispute. Running the USPTO search before purchase is the cheapest filter that keeps the buyer out of the cybersquatting and UDRP mechanisms documented elsewhere in this hub.
Clear the name, then own the domain that holds up
A USPTO search is one half of buying a domain well. The other half is sourcing a name whose history and authority hold up alongside a clean trademark result. A domain that clears the federal register and carries a real, earned backlink profile is the asset; a name with a register conflict or a junk history is the liability. SEO Domains operates the curated marketplace where both checks happen before a domain is listed.
Why the search and the source go together
The USPTO search clears the legal layer: the name does not collide with a live federal mark in the buyer’s class. The sourcing decision clears the value layer: the domain carries a genuine history and an earned profile, not a toxic one. A name that passes the register but fails the profile check is no bargain, and a name with strong metrics but a live trademark conflict is a dispute waiting to happen. Both layers have to clear for the domain to be a real asset.
How to source a domain that clears both layers
A domain that holds up survives two checks before money changes hands. The trademark check is the USPTO search in this walkthrough. The provenance check reads the name’s registration history through ICANN registration data, its prior use, and its backlink profile. The signals that matter are documented across this hub:
- A clean federal register result for the brand string in the domain’s intended class.
- A real registration history, read through ICANN registration records, with no abuse pattern.
- An earned, editorially genuine backlink profile, not a spam-inflated one.
- Prior topical use that aligns with the domain’s planned direction.
A domain that fails the register check carries a legal string. A domain that fails the provenance check carries a value problem. A domain that clears both is an asset whatever a buyer builds on it.
Browse aged and expired domains with the name screened first
The practical demand behind a USPTO search is a domain a buyer can own and build on without a trademark conflict waiting underneath it. That is the product: a screened domain, not a trademark-search service and not a legal opinion. SEO Domains operates the curated marketplace where aged and expired domains are screened across their registration history and backlink profile before they are listed, so the brand string starts from a shorter, cleaner due-diligence list.
