How long does an Indian company actually last?
Of the 14,75,643 companies and LLPs registered in India before 2015, 7,31,186 have since been struck off. Survival rates by year of incorporation and by state, from the register itself.
There are 37,60,338 companies and LLPs on India’s register, and about a quarter of them have been struck off. That much has been written about before, including here.
It is also the least interesting way to read the number, because it mixes a company incorporated in 1989 with one incorporated last Tuesday. The second one has not had time to fail.
Ask the question the other way round. Take every company registered in a single year, and ask what has become of that year’s intake since. The register answers that, and the answer is worse than a quarter.
Almost half of everything registered before 2015 is gone
Of the 14,75,643 companies and LLPs registered before 2015, 7,31,186 have been struck off. That is 49.6%.
Gone from the register, which is not the same thing as having failed in business. Hold that distinction, because further down it turns out to explain most of the number.
As the company-status guide explains, strike-off is not dormancy. A company with that status has been removed from the register and stands dissolved, whatever its website, its letterhead or its GST number still says. What does not disappear with it is the money: section 248 keeps the company’s assets available for its liabilities, and the liability of its directors and officers continues as if it had never been dissolved.
The rate by year of incorporation:
| Registered in | Registered | Still active | Struck off |
|---|---|---|---|
| 1989 | 22,085 | 7,350 | 58.8% |
| 1995 | 59,405 | 21,390 | 55.1% |
| 2000 | 32,376 | 11,596 | 54.4% |
| 2008 | 70,329 | 34,875 | 42.0% |
| 2014 | 81,476 | 41,629 | 45.2% |
| 2020 | 1,80,271 | 1,45,040 | 18.5% |
| 2024 | 2,38,712 | 2,33,458 | 1.7% |
The peak is 1989, at 58.8%, and the years around it sit within a point or two. This is a plateau across the late 1980s and early 1990s rather than one bad year.
The recent years are not good news
The obvious reading of that table is that companies registered recently are doing better. They are not. They are younger.
Removal is not a penalty for a late return. Section 248 of the Companies Act 2013 lets the registrar strike a company off when it has not carried on any business or operation for the two immediately preceding financial years and has not applied for dormant status, or when it never commenced business within a year of incorporation. A long filing gap is usually the evidence of that rather than the trigger itself, and the registrar still has to act on it. A company incorporated in 2024 has had time for neither.
So the curve after about 2015 is measuring how long removal takes, not how long companies last. Any comparison between a 1995 cohort and a 2020 cohort is a comparison between a finished story and one in its first chapter. The recent rows in the table above belong to that unfinished part of the record.
Most of this is one clean-up drive, not a mortality curve
Here is the part the register itself cannot tell you, because it records what a company is now and never when it changed.
After demonetisation the Ministry of Corporate Affairs ran a special drive to clear dormant companies off the register. Answering a question in the Lok Sabha in January 2018, the Minister of State for Corporate Affairs said that registrars had identified 2.97 lakh companies during 2017-18 that had not filed financial statements or annual returns for two or more consecutive financial years and were prima facie not conducting any business, and had already removed 2,26,166 of them as on 19 December 2017. A later answer in the Rajya Sabha put the drive at 3,82,581 companies over three years, identified the same way.
Our snapshot holds 9,49,805 entities at Strike Off. On the government’s own numbers, roughly two in five of them were removed in that single three-year drive.
That reframes the 49.6%. A company incorporated in 1993 that quietly stopped trading in 1999 sat on the register as Active for eighteen years and was removed in 2018. In this data it is indistinguishable from one that collapsed last year. The strike-off count is a record of when the registrar caught up, not of when businesses died.
Two things follow, and both cut against the easy reading:
- The real failure rate for the older cohorts is higher than 49.6%, not lower. Companies that stopped trading before the sweep but were not caught by it are still sitting at Active.
- The timing of removal is a policy decision. A different government with a different appetite for register hygiene would produce a different curve from the same underlying businesses.
Voluntary departure is a rounding error by comparison. The Easy Exit Scheme ran twice, in 2010 and 2011, and 35,174 companies used it; Fast Track Exit had been used by 11,623 more by the end of 2013. Around 47,000 companies chose to leave, against 3,82,581 shown the door.
One oddity we are not going to explain away: the recent years are not smooth. The 2017 cohort shows 28.4% struck off while 2018 shows 11.6% and 2016 shows 18.9%. The obvious candidate is the ground now at section 248(1)(d), for subscribers who never paid up, but that was inserted with effect from November 2018 and does not cleanly reach companies incorporated in 2017. We would rather leave it open than reach for the nearest plausible mechanism.
Where you register matters, or at least correlates
Among states with at least 20,000 entities on the register, the share struck off runs from 42.5% to 16.4%.
| State | On the register | Struck off |
|---|---|---|
| Chandigarh | 20,752 | 42.5% |
| Delhi | 5,10,590 | 32.4% |
| Tamil Nadu | 2,54,855 | 32.2% |
| Punjab | 57,763 | 31.1% |
| West Bengal | 2,94,273 | 27.8% |
| Chhattisgarh | 21,465 | 19.4% |
| Haryana | 1,35,195 | 17.7% |
| Bihar | 82,238 | 17.4% |
| Uttarakhand | 24,201 | 16.4% |
Read this carefully, because it is the easiest table on the page to misinterpret. A state’s share depends on when its companies were registered. Somewhere that saw most of its incorporations in the last five years will look healthy whatever happens to those companies, and somewhere with a long tail of 1990s registrations will look poor. Delhi and Tamil Nadu have deep back catalogues. Uttarakhand, formed in 2000, does not.
The share is a fact about each state’s register. It is not a ranking of where companies survive, and nothing here establishes a cause.
The same figures, one level down, are on the district pages: what each district holds, and how much of its pre-2015 intake has gone.
What this changes for you
If you are checking a counterparty, the base rate is worth carrying in your head. A company incorporated in the 1990s and still Active today is in the surviving half of its cohort, and it survived a registrar actively looking for companies like it. That is genuinely informative, and it is free to check.
The separate industry analysis asks whether those rates also differ by the activity code in a company’s CIN.
What the base rate cannot do is tell you about the company in front of you. Status is a legal state, not a health check: Active means the registrar has not removed it, not that it trades, pays its suppliers, or has filed anything this decade. A company can sit at Active for years while quietly not filing, which is the state that precedes strike-off rather than the opposite of it. The annual filing record is where that shows up, and it is in the filings rather than the register summary. What the filings hold is the next question, and it is a different one from anything a base rate can answer.
Where these numbers come from
Every figure above is a count of the Ministry of Corporate Affairs company
master data, taken state by state from the Ministry’s own register on
18 September 2026 and stated there as on 6 August 2026. No sampling, no modelling, no third-party database:
37,60,338 rows, counted. The calculation is reproducible in
scripts/register_survival.py.
The strike-off drive figures are not ours. They are what ministers told Parliament, in the two written replies linked above, and they count companies where our column counts companies and LLPs together. Removals later set aside by the Tribunal are netted out of neither.
Two limits travel with the snapshot. It is one dated moment, so a company struck off the week after still reads Active here. And entities with an unusable incorporation date, 2,296 of them, are excluded from the year tables and included in the register total.
An earlier version of this analysis counted a different published copy of the same register, and the Chhattisgarh row above is the reason it does not any more. That copy omits Chhattisgarh’s companies altogether, carrying a single entity with a Chhattisgarh CIN where the Ministry’s own figures give 21,465, registered between 1934 and 2026. They are not filed under a neighbouring state either: Madhya Pradesh differs between the two by about two thousand, the ordinary gap between two dates, not by twenty-one thousand reassigned companies. The lesson generalises past this one state. A state table is only as complete as the copy underneath it, and a copy can agree with its own published total while a whole state is missing from it.
Frequently asked
What share of Indian companies get struck off?
Across the whole register, 9,49,805 of 37,60,338 entities carry the status Strike Off, which is 25.3%. That figure understates the rate for any given cohort, because it includes companies registered too recently to have been removed yet. Of the entities registered before 2015, 7,31,186 of 14,75,643 have been struck off, or 49.6%.
How long does a company survive in India?
The register does not record a lifespan, only a current status, so this cannot be answered directly. What it shows is how much of each year's intake has since been removed: 58.8% of the 1989 cohort, 54.4% of 2000, 42.0% of 2008 and 45.2% of 2014. The older the cohort the more of it has gone, though not smoothly, and cohorts from about 2015 on are too young to compare.
Why have so many Indian companies been struck off?
Mostly one clean-up drive rather than a wave of business failure. After demonetisation the Ministry of Corporate Affairs ran a special drive against dormant companies, identifying them by non-filing of financial statements for two or more consecutive years. Registrars removed 2,26,166 companies by 19 December 2017 and 3,82,581 over three years, which is roughly two in five of every entity now showing Strike Off. Many of those companies had stopped trading years before they were removed, so the count records when the registrar caught up rather than when a business ended.
Which Indian state has the most struck-off companies?
By share, among states with at least 20,000 entities, Chandigarh is highest at 42.5%, followed by Delhi at 32.4% and Tamil Nadu at 32.2%. Uttarakhand is lowest at 16.4%. These are shares of each state's own register, not counts.
