How to Search Cases When the Name Is Misspelled

How-To · Case Search

Updated for 2026 · Claw Editorial

A case search that returns nothing is not always a sign the case does not exist. Often the case is there, but the name is spelled differently in the database than the way you typed it. Indian court records carry names from many languages, transliterated in different ways by different clerks at different times, so a single party can appear under three or four spellings in the same court system. This guide explains why that happens, gives you practical techniques to search around it, and tells you which features in modern tools handle the problem automatically.

01Why names are misspelled in court records

Court records in India are not typed by one system. They are entered by clerks across hundreds of registries, in multiple languages, over many decades. Name inconsistencies enter the record at several points.

Transliteration has no single standard

Most Indian names come from Hindi, Urdu, Tamil, Telugu, Kannada, Malayalam, Marathi, Punjabi, Gujarati, or Bengali. When a clerk types a name in Roman script, there is no single correct spelling. "Krishnamurthy" and "Krishnamoorthy" refer to the same name. "Mohammed", "Muhammad", "Mohd", and "Md" are all used for the same given name. The same person may appear differently in filings from different states, or even different filings in the same court.

Old records were scanned, not re-keyed

Many older judgments were digitised from physical paper using OCR (optical character recognition) software. OCR misreads unclear handwriting and printed characters. A "u" becomes an "n", a "ri" becomes "n", an "S" becomes a "5". These errors are baked into the digital record and are rarely corrected.

Different filings use different forms of the same name

A party may shorten or expand their name between a lower-court filing and a High Court appeal. Married women sometimes appear under different surnames. Companies use abbreviated forms. A case that starts as "Rajagopalachari v State" may be cited elsewhere as "Rajagopalan v State of Tamil Nadu" if the clerk used a shortened or alternate form.

Database entry compounds the problem

When legal databases index judgments, they often copy the name as it appears in the document header. If that header already had a typo, the database inherits it. Correcting individual judgment headers at scale is not practical, so the error persists.

02Before you search: prepare the name

Spending two minutes on name preparation before you open a search tool saves much more time in failed searches.

• Write out all spellings you know. If you have the original filing, look at how the name is written there. Also check any correspondence, power of attorney, or cause list entries you have access to.
• Identify the root sound. Strip suffixes and prefixes: "Subramaniam", "Subramanian", "Subramanyan", and "Subramanyam" all derive from the same root. Searching for the shorter form "Subramani" will catch all of them if the tool supports partial matching.
• List common alternate transliterations. "Venkatesh" and "Venkatesh" look the same but may also appear as "Venkatesa", "Venkatesha", "Venkateshwar", or "Venkat". Urdu-origin names like "Farooq" may appear as "Faruq", "Faruk", or "Farooqi".
• Note company name variants. If the party is a company, write down its full registered name, its common short name, and any known older name if it was renamed.

A failed search is rarely proof that the case does not exist. Usually the case is there under a different spelling, and the right technique finds it.

03Technique 1: Try common alternate spellings

The simplest technique is also the one most people skip: run the search two or three more times with alternate spellings before concluding nothing exists.

For each name you are searching, generate two or three plausible alternates and search each one. Common patterns to try:

• "ai" vs "ay" (Pillai / Pillay / Pillae)
• "a" vs "ah" at word end (Varma / Varmah)
• "oo" vs "u" (Gopal / Gopalu; Moorthy / Murthy)
• "ee" vs "i" (Nair / Naier; Suri / Suree)
• "sh" vs "s" (Sheth / Seth; Shankar / Sankar)
• "ph" vs "f" (Farooq / Pharooq, rare but occurs)
• Prefix "Sri" or "Shri" added or removed
• "Mohammad" short forms: Mohd, Md, Muhammad, Mohammed

Each search takes seconds. Running four or five variants is faster than trying to diagnose why you got no results.

04Technique 2: Search by case number or CNR

If you have the case number, CNR number, or year and court, use those instead of the name entirely. A case number does not suffer from spelling variation.

The CNR (Case Number Record) is a unique identifier assigned to every case filed in district and subordinate courts through the eCourts system. It remains constant through the life of the matter. If you can obtain the CNR from any document in the matter, it gives you a direct lookup that bypasses the name problem.

For High Courts and the Supreme Court, the case diary number or SLP/W.P. number similarly identifies the matter precisely. Search using the number, the year, and the court together if the tool supports it.

If you only have a partial case number, try searching for the number without the year, or with the year as a filter rather than part of the query string.

05Technique 3: Use truncation and partial names

Most legal search tools support truncation or partial-name search. Instead of typing the full name, type only the first seven or eight letters and let the tool return everything that starts with that string.

For example, instead of "Krishnamurthy", search "Krishna". You will get more results, but you can then filter by court, year, or subject to narrow down. The name you want is almost certainly in that wider set.

For multi-word names, try searching only the least common part. "Rajan" is common; "Viswanatha" is less common. Searching "Viswanatha" alone will return a smaller, more manageable set than searching the full name, and it is less sensitive to how the less-distinctive part was spelled.

For company names, drop articles and common words. Instead of "The Board of Trustees of the Port of Mumbai", search "Port of Mumbai" or just "Port Trust Mumbai".

06Technique 4: Use phonetic or AI-based search

Phonetic search is the most direct technical solution to the misspelling problem. It works by converting a name to a sound-code (or a learned representation) so that names that sound the same match each other, regardless of how they are spelled.

Classic phonetic algorithms like Soundex were designed for English names and work poorly for Indian names. Modern approaches either use India-specific phonetic rules or train a model on a large corpus of Indian names to learn which variants are equivalent.

The practical effect is that a phonetic-enabled search for "Krishnamurthy" will also return results for "Krishnamoorthy", "Krishnamurti", and "Krishnamoorthi" without you listing each variant. For Indian names with many transliteration variants, this can be the difference between finding a case and missing it.

AI-based semantic search can also help indirectly. If you describe the dispute in natural language rather than typing a party name, the search may surface relevant cases even if the name is wrong, because the legal issue and facts still match.

07Technique 5: Search by judge, date, or subject

When name-based searches fail, pivot to other facts you know about the case. Most legal databases let you filter or search by combinations of fields.

• Judge name: If you know the bench that decided the case, search for judgments by that judge in that court in the relevant year. The results set will be much smaller and the case you want will be easier to spot.
• Date or year range: Narrow to the year or quarter in which you believe the judgment was delivered. Combined with a partial name, a date filter cuts results drastically.
• Subject or statute: Filter by the act or section at issue, for example "Section 138 Negotiable Instruments Act" or "Order 39 Rule 1 CPC". Combined with the court, this can reduce the search to a few hundred results where you can scan for the party name visually.
• Counsel name: If you know which advocate appeared, searching by counsel name is often more reliable than party name, because counsel names are usually typed consistently by the same registry.

Use these facets in combination. Two unreliable fields together often give a reliable result.

08When nothing works: offline and registry options

If digital search across all techniques has failed, the case record may exist only in physical form, or it may be indexed in a way that no standard search reaches it.

• eCourts portal: The eCourts services portal lets you search by party name, case number, act, or filing date across enrolled courts. Try the party-name search with the widest possible name fragment and the correct court.
• Court registry directly: For older or district-level matters, the fastest path may be to contact the court registry. Provide the approximate year, the names of both parties, and the approximate subject. A registry clerk can often pull the physical file number in minutes.
• Cause lists and daily orders: If you believe the matter is active, look at the cause list for the court on the dates when you expect hearings. Cause lists often carry party names alongside case numbers, giving you the number even if the name is garbled.
• Advocate on record: If you know the firm or advocate who appeared, ask them for the file number directly. This bypasses the name problem entirely.

For matters involving contract disputes, litigation funders and corporate legal teams face a similar name-matching challenge at scale. The approach is the same but applied to portfolios rather than single cases. See contract management and legal tools for context on how corporate legal teams organise matter data to avoid this problem in the first place.

09Where Claw fits

Claw is an all-in-one legaltech platform for Indian advocates, law firms, and corporate legal teams, combining AI-based case search, an AI legal assistant (Legal GPT), case management, and compliance automation across all Indian courts and tribunals.

For the misspelled-name problem specifically, Claw builds name-tolerant search into its case search layer. Its AI-based search is designed to handle proximity matching and phonetic variants for Indian names, so a search for "Krishnamurthy" surfaces results recorded as "Krishnamoorthy" or "Krishnamurti" without you needing to run each spelling separately. The database covers 30 crore judgements across 25 High Courts (1980 to 2026) and the Supreme Court (1950 to 2026), with results returned in under 5 seconds. Citations are verified and court-ready.

If you are looking at this problem from a corporate or due-diligence angle, such as screening parties across a large contract portfolio, the case management and compliance features become relevant alongside the search capability.

10Frequently asked questions