,

Same Name, Wrong Person: Namesakes in Indian Court Record Checks

Common Indian names return lakhs of court records. See why a name match is not an identity, and how to read confirmed, probable and excluded results.

·

·

Same name, wrong person: why name-only court record checks in India flag the wrong person, with September 2026 namesake counts for Rahul Sharma and Anil Kumar

A name search finds records. It takes evidence to find a person.

Finding court records for a name is easy. Deciding which of them belong to your person is the hard part. Indian court case records have no Aadhaar or PAN field, and in our September 2026 search the name Rahul Sharma appeared as a party in 31,000+ court records. A fair check separates confirmed, probable, possible and excluded matches, and never treats a shared name as a finding.

A false positive in a court record check is a case that belongs to somebody else but gets attributed to your subject because the names match. This post measures how big that risk is for common Indian names, using live data, and shows how to read a report so that a namesake never costs someone a job, a loan or a flat.

Start with your own name. No Aadhaar or PAN upload. Every finding links to the court record behind it.

Last updated: 3 November 2026

Key takeaways

  • A name is not an identity. In our September 2026 exact-phrase search, Rahul Sharma appeared as a party in 31,000+ Indian court records, Ramesh Kumar in 3 lakh+ and Anil Kumar in 8 lakh+.
  • One name, many spellings. Mohammad Salim, Mohd Salim, Md Salim and Mohammed Salim together return about 40,000 records, and each spelling clusters in different states.
  • Father’s name, age and city are corroboration, not identity. In our published worked example, two men one letter apart shared a father’s name and a birth year, but one lived in Ghatkopar and the other in Jalgaon.
  • Confirmed, probable, possible and excluded mean different things. Only confirmed records should be read as the subject’s. For a person, CLEAR means nothing could be confirmed, not that no case exists; for a company, UNRESOLVED with a score of 0 is not a clean report.
  • Role matters as much as identity. Someone who filed a case and someone a case was filed against are not the same risk.

How many people share a name in Indian court records?

Enough that a name on its own identifies nobody. On 25 September 2026 we ran an exact-phrase litigant search for 16 common two-word Indian names across the index of eCourtsIndia (ecourtsindia.com), an independent, non-government platform holding 32 crore+ (320 million+) court and tribunal records. The smallest result was about 3,000 records. The largest was over 8 lakh.

Each figure is the number of court records in which that name appears as a party. It is not a count of people. One man with ten cases is ten records, and ten thousand different men called Anil Kumar are ten thousand people. A name search cannot tell the two apart. For everything else a court check covers beyond this identity problem, see what a court record check covers.

Name searchedCourt records naming the litigantTop state (share of records)Next two states
Anil Kumar8,14,413Uttar Pradesh (31%)Bihar, Haryana
Manoj Kumar6,80,641Uttar Pradesh (33%)Bihar, Rajasthan
Rajesh Kumar6,42,935Uttar Pradesh (30%)Bihar, Rajasthan
Santosh Kumar4,63,197Uttar Pradesh (34%)Bihar, Madhya Pradesh
Amit Kumar4,13,277Uttar Pradesh (35%)Bihar, Haryana
Ravi Kumar3,08,560Uttar Pradesh (20%)Bihar, Andhra Pradesh
Ramesh Kumar3,04,598Uttar Pradesh (17%)Rajasthan, Haryana
Sunita Devi1,33,605Bihar (37%)Uttar Pradesh, Jharkhand
Sanjay Singh95,173Uttar Pradesh (36%)Bihar, Madhya Pradesh
Abdul Rahman46,660Uttar Pradesh (21%)Kerala, Tamil Nadu
Suresh Yadav35,311Bihar (36%)Uttar Pradesh, Jharkhand
Rahul Sharma31,276Uttar Pradesh (28%)Rajasthan, Madhya Pradesh
Ashok Patil17,786Maharashtra (83%)Karnataka, Gujarat
Mohammad Salim13,181Maharashtra (38%)Rajasthan, Uttar Pradesh
Joseph Thomas3,432Kerala (84%)Maharashtra, Karnataka
Priya Sharma3,094Uttar Pradesh (22%)Punjab, Rajasthan
Lakshmi*6,98,516Andhra Pradesh (40%)Tamil Nadu, Karnataka
Venkatesh*4,12,841Karnataka (58%)Telangana, Andhra Pradesh
Murugan*2,92,717Tamil Nadu (90%)Karnataka, Kerala
The Namesake table. Exact-phrase litigant search across all courts and tribunals in the eCourtsIndia index, read live on 25 September 2026. Shares are the top state’s records as a share of the total. *Single names match that word anywhere in a party string, including inside a relative’s name, so they are not directly comparable. Counts are records, not people.

Where we also pulled the court-level split, most of the crowd sits in district and taluka courts. Anil Kumar: 6,80,007 district and taluka, 1,13,703 High Court, 13,633 tribunal and 7,070 Supreme Court records. Ramesh Kumar: 2,50,598, 46,764, 4,387 and 2,865. Sunita Devi: 1,20,704, 11,236, 1,061 and 605. Rahul Sharma: 27,673, 2,910, 451 and 242.

Anyone can rerun these searches on the free eCourtsIndia litigant search, and our litigant search guide shows how.

What the table tells a reviewer

  • “Kumar” does a lot of the damage. Across much of north India, Kumar works as a middle name or a stand-in surname, so a phrase search for Anil Kumar also returns every longer name that contains it. Our launch essay counted 2.5 crore+ records naming a litigant matching Kumar alone (14 September 2026).
  • Names are regional, and so are namesakes. 90% of Murugan records sit in Tamil Nadu, 84% of Joseph Thomas records in Kerala and 83% of Ashok Patil records in Maharashtra. For a subject from those states, a state filter barely thins the crowd.
  • A state filter helps less than people expect. Rajesh Kumar returned 1.9 lakh+ records in Uttar Pradesh alone. A district narrows it further, but geography can mislead too.
Bar chart: court records naming 16 common Indian names as a party, from Anil Kumar (8 lakh+) to Priya Sharma (3,000+), eCourtsIndia search, 25 September 2026
How many court records share one name: exact-phrase litigant counts from the eCourtsIndia index, read live on 25 September 2026.

One caution if you rerun these. Plain counts for Ramesh Kumar and Sunita Devi came back at almost exactly 1 lakh earlier in the day, while facet-backed reads of the same searches returned the larger totals above. A count sitting near 1 lakh for a very common name is a floor, not a total. Sources and method has the detail.

What a full-text search adds (October 2026)

The table above counts only records where the name is a party. Many background-check vendors search more widely, running the name through the whole record, where it also matches advocates, judges and the text of orders. Re-run that way on 12 October 2026, “Rahul Sharma” returned 80,000+ records in the eCourtsIndia index, about two and a half times the party-only count, and “Anil Kumar” returned 19 lakh+. The wider the field a vendor searches, the bigger the crowd it hands your reviewer.

Those 80,000+ Rahul Sharma records are also a reminder that a count is not a risk profile. About 53,000 are disposed and 19,000 pending. The single biggest court in the set is the Supreme Court, with 10,000+ records, followed by the Rajasthan High Court at Jaipur and the Punjab and Haryana High Court. None of that tells you anything about any one Rahul Sharma. You can re-run the search and watch the counts change as you add a state or a filing year.

Why is a name not an identity in India?

Because Indian court records were built to list parties, not to identify them. There is no Aadhaar or PAN in a cause title, and the name itself changes shape from one register to the next.

Spelling and transliteration

Court registers are typed by hand, often from a name first written in another script, and every court makes its own choices. Live counts for seven spellings of two names, read on 25 September 2026:

Spelling searchedCourt recordsWhere it clusters
Mohammad Salim13,181Maharashtra (38%), then Rajasthan and Uttar Pradesh
Mohd Salim12,823Maharashtra (43%), then Uttar Pradesh and Delhi
Md Salim8,877Bihar (45%), then West Bengal and Jharkhand
Mohammed Salim5,124Maharashtra (44%), then Karnataka and Rajasthan
Mohammed Rafi7,421Karnataka (41%), then Kerala and Tamil Nadu
Mohd Rafi6,419Uttar Pradesh (50%), then Jammu and Kashmir
Mohammad Rafi6,239Andhra Pradesh (30%), then Uttar Pradesh and Karnataka
Exact-phrase litigant counts for spelling variants, eCourtsIndia index, read live on 25 September 2026. Variants can overlap where one record carries two spellings.

A check that searches one spelling misses the rest: the four Salim spellings together come to about 40,000 records before overlap, and the most common one is only about a third of that. Spelling is also a regional habit. In this data “Md” is the Bihar, Bengal and Jharkhand form, while “Mohd” shows up strongly in Uttar Pradesh, Delhi and Jammu and Kashmir. A subject from Patna whose form says Mohammad may well be recorded as Md.

This is not a feature of one community’s names. Our LegalCheck launch essay found one evaluated subject’s given name spelt six ways: Siddharth, Siddharath, Siddhrath, Sidharath, Sidhdarth and Sioddharth. The worked example in our LegalCheck API guide found one man written three ways across nine matters, two of them with a nickname inside the party field.

Initials and honorifics

Southern registers often write a person as an initial and a given name, the initial standing for the father or family. A name with a single-letter token looks statistically rare. It identifies nobody. Northern registers add titles: Shri, Sh., Sri, Smt and Late.

On the first page of results for Rahul Sharma, a petition described a woman as the wife of “Shri Rahul Sharma”. Treat Shri as a first name and you find agreement that is not there, which is the third trap below.

Patronymics, surnames, and names that belong to someone else

Many Indian names carry a relative inside them. In Maharashtra a married woman is often recorded with her husband’s full name after her own, so a phrase search for Rahul Sharma also finds women whose husbands are called Rahul Sharma. On the first page of 40 results for that name, in relevance order on 25 September 2026, at least three records carried the phrase only inside a woman’s name, and several more carried it after “W/o”, wife of. In none of those three is a man called Rahul Sharma a party in his own right.

Surnames add a second layer. Kumar, Devi, Singh and Yadav are shared by very large populations, so sharing a surname with the subject’s father says almost nothing. As the launch essay puts it, a shared surname is not a family. The API guide notes that a father’s name is compared by the stem of the given name, not by string equality, and the published gate list includes one named score/patronymic-stem.

Mass-party cause titles

A cause title can list dozens or hundreds of parties, so three search words can come from three different people. The second trap below shows it happening in a real check.

No identity number to fall back on

A father’s name is missing on a large share of civil and commercial filings, and addresses are free text when they exist at all. Identity has to be built from evidence, record by record, which is why the structure of the underlying data matters as much as its size.

Are father’s name, age and city enough?

No. They are corroboration, not identity. They make a match more or less likely. On their own, they do not prove that a record belongs to your person.

Almost every court check in India asks for these fields, and they are the right ones, because they are the only ones Indian court metadata gives anything to match on. The mistake is treating agreement on them as proof.

The two men in our worked example

Our API guide ran one real subject live on 15 September 2026 and published every step. A name that returned 63 court records nationally resolved to 9 records covering 8 distinct disputes. Before any court search, the electoral roll found two men whose names differ by a single letter. Both were male. Both were recorded at age 33, so both were born around 1991. Both had a father with the same name. Both scored 100 out of 100 on name and relative.

One lived in Ghatkopar in Mumbai. The other lived in Jalgaon, in north Maharashtra. Only one of them had the cases.

A system that ranks candidates by similarity and takes the top one had roughly even odds of choosing the wrong man. What separated them was the submitted address, compared on letters rather than strict equality, and then a corroboration no name search can reach: the police station prosecuting several of the matters served the same neighbourhood as the Ghatkopar man’s polling area.

When a perfect match is the wrong person

The launch essay records two more cases. During development, a Jalgaon matter carrying IPC 307 (attempt to murder, now section 109 of the Bharatiya Nyaya Sanhita) matched a subject on name and on father’s name, exactly. It belonged to a different man. Separately, an elector who matched 90% on name and relative was withheld, because 14 people in the country carried that exact pair and none of them lived in Howrah, where the subject’s address resolved. Both refusals are printed in the report, with reasons.

So give a check every identifier you legitimately hold, and then read what it did with them. In a LegalCheck report, each match shows whether the name, father’s name, address, date of birth and PAN agreed in full, in part, closely, or were not provided (developers will find this in the match_breakdown field).

What do confirmed, probable, possible and excluded mean in a background check?

They are levels of identity confidence, not levels of guilt. Confirmed, probable, possible and excluded describe individual records. CLEAR and UNRESOLVED are bands for the report as a whole, explained in the next section.

LabelWhat it meansWhat the reviewer should doWhat the reviewer must not do
ConfirmedThe evidence supports attribution strongly enough to present the matter as this person’s, with an evidence chain and a CNR you can open.Open the CNR, read the orders, check the person’s role, then give them a chance to explain.Treat it as proof of wrongdoing. A pending case is an allegation, and disposed is a status, not an outcome.
ProbableMeaningful evidence, but an identity question remains.Review the match breakdown beside the cause title. Ask the person, or seek one more identifier.Merge it into confirmed, or take adverse action on it alone.
PossibleSurfaced for awareness. Never counted as an identity match.Note it. Look closer only if something else points to it.Count it, score it, or describe it to anyone as a finding.
ExcludedThe record was found and considered, and attribution was rejected, with the reason captured.Read the reason, especially on serious-looking matters, and keep it on file as evidence of diligence.Pull it back in because the charge looks alarming.
UNRESOLVED (companies and firms only)Many records sit under the company’s name and none could be tied to this particular entity. Its score is 0 only because the score counts verified matters.Route it to manual review. Resubmit with the exact registered name.Map it to green, clear or pass.
CoverageWhat was searched, what could not be checked, and where the boundaries are.Read it before relying on a clear result.Assume anything outside it was checked.
Label meanings as published in the LegalCheck launch essay (14 September 2026) and the LegalCheck API guide (15 September 2026). UNRESOLVED is a report-level band set only for companies and firms (LegalCheck report rules, read 12 October 2026), not a fifth kind of match.

A vendor that handed a reviewer all 63 name matches in the worked example would have been attributing 54 other people’s records to one man. Ask for confirmed matches and records screened as two separate numbers, never a single “cases found” total.

Why is a score of zero not a clean report?

Because zero can mean two opposite things: nothing was found, or plenty was found and none of it could be tied to your person. Only the label tells you which.

For a person, a LegalCheck report reads CLEAR when nothing could be confirmed, even if two hundred same-name records sit underneath as probable or possible. That is correct, because none of them has been shown to be your person, but it does not mean the person has no cases. Read the candidate list and the coverage note before you treat CLEAR as clean.

For a company or firm there is a separate band, UNRESOLVED, which carries a score of 0. It means many records sit under the name and none could be tied to this particular entity, usually because the registered form of the name was not given or several entities share it. Almost every dashboard paints zero green; UNRESOLVED is not clear. Route it to manual review and resubmit with the exact registered name.

Two related rules: the launch essay treats a count of zero from a name search as a broken query, not a rare name. And when a very common name has nothing to separate it, a job can end in a MAX_RESTARTS failure: the engine declines to guess rather than return a plausible stranger, and a failed check is not charged.

A clear report is not a clearance. It means nothing could be confirmed within the stated coverage and the evidence available. It does not prove that no case exists anywhere, and it says nothing about an FIR that has not reached a court. Our comparison of court checks, police verification and FIR search explains which check answers which question.

What are the three traps that inflate false positives?

Counting one dispute twice, trusting a crowded cause title, and reading a title as a name. Each has a named rule, a gate, in the published LegalCheck model, and each gate can only lower a match, never raise it. The gate names below come from our LegalCheck API guide.

1. The same dispute counted twice

Gates: score/same-dispute-is-one-source and score/one-dispute-counts-once. In the worked example, a bail application in the City Sessions Court and the sessions case it belonged to carried substantially the same sections, the same co-accused and the same police station. Counting both would have doubled the apparent criminal history of a man arrested once. A suit and its execution, or a petition and its appeal, are the same pattern.

For reviewers: count disputes, not rows. Ten rows on a report can be three disputes.

2. Mass-party cause titles

Gate: score/mass-party-cannot-identify. Of 12 Mumbai records in the worked example, three were mass-party titles: one listing roughly fifty respondents, one about a hundred and forty, and one listing eleven, where the subject’s given name, his father’s given name and his surname were each contributed by a different man. All three were removed. The launch essay also describes a Supreme Court matter with several hundred parties that was held at possible and never promoted.

For reviewers: if your subject’s name is one of fifty on the cause title, that record cannot identify your subject by itself.

3. An honorific read as a first name

Gate: score/honorific-not-a-given-name. The launch essay explains the failure behind it. Sh., Shri, Sri and Late are titles, and before this constraint existed, an honorific on the form could match an honorific in the record and mint a father out of nothing. “Late” deserves special care: it often sits in front of a father’s name, and it tells you something about the father, not about the subject.

For reviewers: strip titles before you compare names, on both the form and the record.

Does it matter who filed the case?

It changes the meaning of the whole report. A person who filed forty recovery suits and a person who had forty filed against them are different risks, and a count-only check calls both “40 cases found”.

Each LegalCheck match records the subject’s role: respondent or defendant, petitioner or complainant, applicant who is the accused (a bail or quashing applicant), both, or unknown. The risk score weighs which side of the case the subject is on, along with the severity and status of the matter. Only confirmed matches move the score; probable and possible matches never do.

Role also interacts with the namesake problem. On the first page of Rahul Sharma results, the name appeared on the prosecution side in “State through” titles, on both sides of family disputes, and as the defendant in bank recovery suits. A reviewer who sees only the name cannot tell a complainant from an accused. People appear in court records as victims, complainants and petitioners every day. Our case status dictionary explains what the status on each of those records does and does not tell you.

What should “accuracy” mean for a court check?

Two numbers, not one: how often a reported match really is your person (precision), and how many of your person’s real cases the check finds (recall).

A precision failure puts a stranger’s case on your subject. A recall failure misses a real case. On the page they look identical. They are not equally harmful: a false match can cost someone a job over a case that was never theirs. LegalCheck is built on that asymmetry. The launch essay says recall improvements have been specified, costed and declined where each could introduce a false match.

Our evaluation, with its limits

MeasureResult
Real subjects scored by a separate evaluator20
Overall score9.8 / 10
False court confirmations0
False electoral identifications0
Missed court matters1, disclosed
Missed electoral identifications2, disclosed as deliberate trades
eCourtsIndia’s own evaluation of the production engine on a twenty-subject evaluation set, scored by a system kept separate from the engine that never sees its reasoning, 14 September 2026. It is not a third-party certification.

Twenty subjects is a real measurement, and a small one: enough to say the engine did not invent matches on that set, not enough to print a precision percentage. So we will not turn 9.8 into a “99%”.

Some vendors do publish one figure. AuthBridge’s court record check page, for example, shows “99% Verification Accuracy” (read 12 October 2026). We could not see on that page whether the figure measures precision, recall, or agreement after its manual quality check, and it may be defined elsewhere. Ask that of any single accuracy number, ours included.

Ask any provider: what is the denominator, how many false confirmations on how many real subjects, are excluded records shown with reasons, and what does an unresolved or clear result look like? Our court record check vendor checklist turns these into RFP questions.

What Indian courts have said about same-name mistakes

Courts have treated a same-name mix-up as a real harm, not a clerical slip. The reported cases are about arrests and credit files rather than background checks, but the lesson carries over: a matching name, even a matching father’s name, does not prove you have the right person.

  • Karnataka High Court, July 2022. A man was arrested on a warrant meant for a former director of a company in liquidation, and the only link was that his father’s name was similar to the one in the warrant. He disputed that he was the man named in the warrant, but his identity was never cross-checked. Justice Suraj Govindaraj held that the arrest violated Article 21, ordered the State to pay him ₹5 lakh (recoverable from the arresting officers), and told the Director General of Police to issue identity-verification guidelines if none existed. Identity comes first, the court said; a father’s name “may have a corroborative role”. Ningaraju N v. Official Liquidator of India Holiday (Pvt) Ltd, Company Application 96 of 2022, decided 7 July 2022, CNR KAHC010295432022 (SCC Times report).
  • Allahabad High Court, Lucknow Bench, February 2026. Two connected cases, one point: police arrested men who shared a name with the real accused. In the older one, a 2004 fraud case, the man arrested had the same father’s name as the accused, but a different mother’s name, and the warrant did not carry the mother’s name at all. On 9 February 2026 Justice Tej Pratap Tiwari quashed the proceedings against both applicants, held that taking away someone’s liberty on mistaken identity strikes at Article 21, and directed police chiefs in Lucknow and Sitapur to act against the officers responsible. The warrant, the court noted, “does not contain complete identifying particulars, such as the name of the mother”. Mohd. Azeem Idrishi v. State of U.P., Application u/s 482 No. 8263 of 2025 with No. 8685 of 2025, CNR UPHC020752402025, 2026 LiveLaw (AB) 71 (LiveLaw report).
  • Supreme Court, 2025 to 2026. Rajendra Singh Panwar, from Uttarakhand, told the Supreme Court his CIBIL record had carried an adverse score since 2020 although he had taken no loan and made no default. His case was that two other people with the same name had been issued the same PAN, and their defaults were landing on his file, even after he was given a new PAN. The court sought affidavits from the banks. PNB said it had reported no default against him, and CIBIL told the court his records had been corrected, as PTI reported on 5 April 2026. Rajendra Singh Panwar v. Jay Prakash and others, SLP (C) 35978 of 2025, before Justices J.B. Pardiwala and K.V. Viswanathan, CNR SCIN010594362025 (PTI report in Telangana Today).

Two of these are arrests and one is a credit file, not a court record check. Each turns on the failure this post describes: a record was tied to a person because the name, and sometimes the father’s name, matched, and nobody compared the details that would have told them apart. That is why a fair court check treats a father’s name as corroboration, and keeps probable and possible matches out of any decision. These are summaries of what the courts said, not legal advice.

What should you do if a check wrongly matches you?

Ask for the case number, check it yourself, and put the differences in writing.

  1. Ask for the CNR, not a summary. Every court case in India’s eCourts system has a 16-character CNR (tribunal cases use their own formats). A finding without one cannot be checked. Our CNR guide explains how to read it.
  2. Open the case. Search the CNR, or your own name, on eCourtsIndia search. Read the cause title and the orders, not just the status.
  3. Compare the details that separate people: father’s or husband’s name, address and district, age, co-parties, the police station, the advocates. Note every one that does not match you.
  4. Check the role. You may be the complainant, or your name may sit inside someone else’s, as a husband’s name inside a wife’s.
  5. Write to the employer, lender or platform that ran the check. Attach your notes, ask which label the finding carried, and ask them to rerun it with your parent’s name, address and date of birth. A probable or possible record should never have been treated as a match on its own.
  6. Keep copies of everything.

If a case of yours is missing from the index, use Add a Missing Case with its CNR. If a record shown on eCourtsIndia is wrong or sensitive, the FAQs and Case Removal page sets out the policy: requests come through a verified lawyer, with the case details and a reason, and are reviewed in accordance with applicable laws. We cannot promise removal, and we cannot change or delete a court’s own record, which sits with the court.

Check yourself first

The easiest way to see how namesakes affect you is to run a check on your own name before an employer, lender or landlord does. The court record check guide explains what a complete check covers, and our background check guide walks through the manual route. Free name search is on eCourtsIndia litigant search; for a scored report that separates you from your namesakes, run a LegalCheck. Families screening a proposed match can read our guide to a background check before marriage.

What this does not mean

  • A match is not guilt. Being named in a case establishes nothing. A person is presumed innocent unless convicted.
  • Pending is not conviction, and disposed is a status, not an outcome. Only the order tells you how a case ended.
  • A large count for your name says nothing about you. These are records naming the phrase, spread across many different people.
  • The counts are not a ranking of any name, region or community. They reflect how common names are and how many records a state’s courts hold, not anyone’s conduct.
  • A score of 0 is not automatically clean. For a company, UNRESOLVED with a score of 0 is not clear. For anyone, a clear report is not a clearance certificate, and it is not FIR or police verification.
  • Our evaluation is not a certification and not an accuracy percentage. It is twenty real subjects, scored adversarially, with the misses disclosed.

Sources and method

  • Namesake counts. eCourtsIndia index, read live on 25 September 2026 through the eCourtsIndia MCP search_cases tool with the litigants filter and nameMatchMode phrase (an exact, ordered phrase anywhere in a party field), pageSize 1, CNR projection and a stateCode facet for the top five states. Court-level splits for Anil Kumar, Ramesh Kumar, Sunita Devi and Rahul Sharma came from the courtLevel facet on the same day. Counts are records, not people. Facet totals can differ slightly from headline counts.
  • The 1 lakh question. Plain counts read earlier on 25 September 2026 returned 1,00,097 for Ramesh Kumar and 1,00,025 for Sunita Devi. The same searches with a stateCode facet returned 3,04,598 and 1,33,605, and the courtLevel facets for Ramesh Kumar summed to 3,04,614 even in a response whose headline count read 1,00,097. We publish the facet-backed totals and treat any headline count near 1 lakh for a very common name as a floor. Anil Kumar returned 8,14,413 with and without facets.
  • Spelling variants. Same method and date, seven spellings.
  • First-page observations for Rahul Sharma. One page of 40 results in relevance order, read 25 September 2026. Qualitative observations, not a measured rate. We did not estimate how many records carry a relative’s name, because we could not measure it cleanly from truncated party strings. No individual’s cases or CNRs are reproduced.
  • Full-text re-run. eCourtsIndia MCP search_cases with the query “Rahul Sharma” (and “Anil Kumar”) as an exact phrase across the combined record (parties, advocates, judges, order text), with caseStatus and courtCode facets, 12 October 2026. Counts are records, not people.
  • Report bands and subject role. LegalCheck report rules, read 12 October 2026: the rating rests on confirmed matches only; UNRESOLVED is set only for a company or firm; a person with only same-name records reads CLEAR; the subject’s role is recorded on every match.
  • Worked example, gate names and match breakdown. LegalCheck API guide, published 15 September 2026, last updated 23 September 2026, read 25 September 2026.
  • Report labels, precision constraints, the Jalgaon IPC 307 and Howrah examples, the six spellings, the Kumar count and the evaluation. LegalCheck launch essay, 14 September 2026, last updated 23 September 2026, read 25 September 2026.
  • Vendor statements. AuthBridge Court Record Check page (authbridge.com/checks/court-record-check), read 25 September and 12 October 2026. Quoted as published. We have not tested it, and eCourtsIndia is not affiliated with any vendor.
  • Case removal policy. ecourtsindia.com/faqs, read 25 September 2026.
  • Court rulings on same-name mistakes. Each matter was checked on the eCourtsIndia index by CNR on 3 November 2026. Karnataka: the final order of 7 July 2022 (CNR KAHC010295432022), read in full. Allahabad: the judgment delivered on 9 February 2026, read in full from the copy linked by LiveLaw; the eCourtsIndia record for CNR UPHC020752402025 had not yet picked up that final order when we checked. Supreme Court: the record of proceedings of 9 January 2026 (CNR SCIN010594362025), plus the PTI report of 5 April 2026 for the later correction by CIBIL; we did not find a final order. All read 3 November 2026.
  • Price. On the web (ecourtsindia.com/legal-check, read 12 October 2026): Rs 100 plus metered AI usage, first 3 completed web checks free, charged only when a report completes. Through the Partner API or MCP (ecourtsindia.com/api/pricing): ₹99 per accepted check on pay as you go, ₹33 with a subscription, report re-reads free.
  • Index size. 32 crore+ (320 million+) court and tribunal records, the evergreen floor for the eCourtsIndia index.
  • Official source. eCourts Services (ecourts.gov.in), the government portal for case status and the CNR. eCourtsIndia is not affiliated with it.
  • Figures are a dated snapshot of a live system, not a service level commitment. The Rahul Sharma count moved by a few records within hours.

Frequently asked questions

Can a court record check flag the wrong person in India?

Yes. Indian court records have no Aadhaar or PAN field, so a check that matches on name alone will pick up other people’s cases. A common name such as Anil Kumar appeared as a party in 8 lakh+ court records in our September 2026 search. A fair check separates confirmed records, tied to your person by evidence beyond the name, from probable and possible ones that share only the name, and shows the case number behind each.

Is a father’s name match enough to confirm a court record belongs to someone?

No. A father’s name is strong corroboration but not identity. In one worked example, two men whose names differed by one letter had fathers with the same name and were the same age, yet only one had the cases. Address, age, co-parties, the police station and the court district all have to be compared before a record is treated as belonging to a person.

What does a probable match mean in a background check report?

A probable match is a court record that carries the subject’s name and some supporting detail, but where an identity question remains. It is not a finding against the person. A reviewer should open the case, compare the parent’s name, address and age on the record, and ask the candidate before acting. In a LegalCheck report, probable and possible records never move the risk rating.

What should I do if a background check wrongly shows a court case against me?

Ask the employer, lender or verification vendor for the CNR or case number of each finding. Open each case and compare the father’s or husband’s name, address, age, co-parties and your role. Write back with the differences and ask for the check to be rerun with your parent’s name, address and date of birth. Keep copies of everything you send.

Why does my name show so many court cases on eCourtsIndia?

Because a search counts records, not people. A full-text search for Rahul Sharma returned 80,000+ records in the eCourtsIndia index in October 2026, because it also matches advocates, judges and order text, and thousands of different people share the name. A large count says nothing about you. Filter by state and district, then open individual cases to see who the party actually is.

Does a CLEAR court record check mean a person has no cases?

No. CLEAR means nothing could be confirmed as the subject’s within the courts searched and the details supplied. It does not prove that no case exists anywhere, and it says nothing about an FIR that has not reached a court. Read the probable and possible records listed underneath, and the coverage note, before treating a clear report as a clean one.

Can eCourtsIndia delete a court case that is shown against my name?

eCourtsIndia cannot change or delete a court’s own record, which sits with the court. For records displayed on eCourtsIndia, the FAQs and Case Removal page asks for a request through a verified lawyer with the case details and a valid reason, and each request is reviewed in line with applicable law. Removal is not guaranteed, and the court’s original record is unaffected.

Have Indian courts dealt with same-name mistakes?

Yes. In 2022 the Karnataka High Court awarded ₹5 lakh to a man arrested only because his father’s name resembled the one in a warrant. In February 2026 the Allahabad High Court quashed proceedings against two men arrested on mistaken identity. In 2026 CIBIL told the Supreme Court it had corrected the credit file of a man whose PAN had also been issued to two namesakes. You can open the Karnataka case record on eCourtsIndia.

Important: LegalCheck is a court-record matching report based on public records and probabilistic identity matching. It is not an identity certificate, FIR search, police clearance certificate or legal advice. A match is not proof of wrongdoing, and a person is presumed innocent unless convicted. The namesake counts in this post are dated counts of records, not of people, and say nothing about any individual. Verify identity, read the case and give the person an opportunity to explain before acting. Use reports only for a lawful purpose and keep them confidential. The court rulings summarised in this post are for information only; this is not legal advice.

Check yourself first

See what India’s courts hold under your name, and how a report separates you from your namesakes, before anyone else runs the check. On the web, a check costs Rs 100 plus metered AI usage, your first 3 completed checks are free, and you are charged only when a report completes (October 2026; see the LegalCheck page for current rates). Through the Partner API or MCP it is ₹99 per check, or ₹33 with a subscription.

Most reports arrive within a few minutes. Common names can take longer.

eCourtsIndia is a private legal-technology platform. It is not affiliated with, associated with, or endorsed by the Government of India, the Supreme Court of India or its e-Committee, or any court. Official case information is published on ecourts.gov.in. Always verify details against official court records or certified copies. This article is general information, not legal advice. Spotted an error? Write to support@ecourtsindia.com.

Search 35 crore+ Indian court case records, free

One search across the Supreme Court, all 25 High Courts, district courts and 18 tribunal and commission types. Hearing alerts, AI summaries and an API for developers.