Data
Company dissolutions in two state registries
Last updated October 23, 2026
Next update November 2026
Every company that closes properly files a last paper with its state. Two states publish those filings in open files, with the date each company was formed, in a second file for Colorado. We counted the dissolutions in both since January 2025: 110,339 companies, 88,564 in Colorado and 21,775 in Connecticut. This page shows how long they had been in business when they closed, and what kind of entity they were.
Most were young. 61.9% had run less than five years. Still, 22,451 of them, about one in five, had been in business ten years or more. That’s a decade of estimates, job files, email and the CRM, and nobody adds to it once the doors shut.
Dissolutions by registry, year and years in business
| Registry and year | Dissolved | Under 2 yrs | 2–4 | 5–9 | 10–19 | 20+ | Dates out of order |
|---|---|---|---|---|---|---|---|
| Colorado 2025 | 49,871 | 16,404 | 13,946 | 9,136 | 6,963 | 3,408 | 14 |
| Colorado 2026, 17 Oct | 38,693 | 12,963 | 10,795 | 7,354 | 5,026 | 2,548 | 7 |
| Connecticut 2025 | 11,491 | 4,589 | 2,878 | 1,604 | 1,476 | 944 | 0 |
| Connecticut 2026, 18 Oct | 10,284 | 4,216 | 2,544 | 1,438 | 1,241 | 845 | 0 |
| Both registries | 110,339 | 38,172 | 30,163 | 19,532 | 14,706 | 7,745 | 21 |
“2026, 17 Oct” means 1 January to 17 October 2026. Each row adds up to its “Dissolved” cell. The mean years in business were 6.4 in Colorado’s 2025 row, 6.2 in its 2026 row, 6.2 in Connecticut’s 2025 row and 6.1 in its 2026 row.
The share that ran ten years or more
| Registry and year | 10+ yrs | Share 10+ | Share 20+ |
|---|---|---|---|
| Colorado 2025 | 10,371 | 20.8% | 6.8% |
| Colorado 2026, 17 Oct | 7,574 | 19.6% | 6.6% |
| Connecticut 2025 | 2,420 | 21.1% | 8.2% |
| Connecticut 2026, 18 Oct | 2,086 | 20.3% | 8.2% |
| Both registries | 22,451 | 20.3% | 7.0% |
Shares are of each row’s “Dissolved” total, rounded to one decimal.
Dissolutions by entity type, 2025 and 2026 together
| Registry | Entity type | Dissolved |
|---|---|---|
| Colorado | Limited liability companies | 76,081 |
| Colorado | Profit corporations | 11,815 |
| Colorado | Public benefit profit corps | 120 |
| Colorado | Limited partnerships and LLLPs | 487 |
| Colorado | Other forms | 61 |
| Connecticut | Limited liability companies | 21,285 |
| Connecticut | Stock corporations | 476 |
| Connecticut | Other forms | 14 |
What the tables show
The two states look alike. In each, about a fifth of the companies that dissolved had run ten years or more. The share barely moves between 2025 and 2026, from 20.8% to 19.6% in Colorado and from 21.1% to 20.3% in Connecticut.
The young end is where they differ. In Connecticut, 40.4% of the companies that dissolved were under two years old. In Colorado it was 33.2%. Connecticut also keeps more at the old end: 8.2% had run twenty years or more, against 6.7% in Colorado.
The long-running companies are not a rounding error. 7,745 had been in business twenty years or more. 14,706 more had run between ten and nineteen. Each of those closed with a run of paperwork behind it longer than most of the companies on the list ever lived.
The entity column says less than it seems to. Limited liability companies are 85.9% of Colorado’s dissolutions and 97.7% of Connecticut’s. That reflects how companies are formed in both states today, not which ones were large. An LLC can be one person with a truck or a firm of eighty.
Read the rows as companies that filed to close, not companies that stopped trading. A company the state dissolved for a missed report is left out, because it may still be trading. The table also can’t say which companies had 25 or more people. The methodology says why.
Why ten years is the line we drew
Ten years is our cut, not a buyer’s. micro1 publishes no minimum number of years. Its payout page asks “How many years of data do you have?” and adds “Typically the number of years you’ve been in business.” Our page on how many years of records count goes through that question.
What micro1 does say is what it looks for: data that “reflects how work is actually performed within an organization.” Our reading: one year shows how a company works. Ten show how it changed. A contractor that closed after twelve years has twelve years of estimates, revised estimates and the change orders where the price moved. A repair shop has its repair orders and the notes on every comeback. A buyer licenses records like these to train AI models.
The records are de-identified before use. Customer names and payments are never what’s wanted. What’s wanted is how the work was done: how an estimate became a job, how a change order was settled, how a repair was diagnosed. Our explainer on what data buyers actually want from a trade business goes through it record by record.
Size still matters. micro1’s referral page lists “Company requirements: 30+ employees”. We refer a closed company to micro1 when it had 25 or more people at its peak. micro1 decides at application. Our page for closed companies covers what the founder keeps and how the steps run. micro1’s data partnerships page says “Companies retain ownership of their underlying data.”
Some of the larger companies that close also file a federal notice first. Our table of WARN notices by state counts those, by the workers each notice names.
If your company is one of the ten-year closures, or about to be, our estimate asks five questions and gives you a band. It’s ours, not an offer. How many of those years can you still open?
What is never included
On this page: counts only. No company names, no addresses, no agent names, no filing text. The states publish those; we don’t repeat them.
In a licence: micro1 says the “Anonymization and redaction requirements” are agreed with each company before any participation. Our advice: keep customer payment details, bank exports, payroll, employee medical records and anything a client owns out of any export. How it’s safe sets out what leaves your systems and when.
Methodology
What is counted. Each number counts dissolution records in our database, read from the two state registers with a read-only query on 23 October 2026.
The window. Colorado: dissolutions received by the Department of State from 1 January 2025 to 17 October 2026, from its business entity transaction history. Connecticut: dissolutions dated 1 January 2025 to 18 October 2026, from the Secretary of the State’s business registry master. The year columns use those dates.
Years in business. The dissolution date minus the formation or registration date the registry gives, in years. For Colorado, the formation date comes from the business entities file; for Connecticut, from the registration date in the business master. The bands are under 2, 2 to 4, 5 to 9, 10 to 19, and 20 or more. A handful of records give a formation date after the dissolution date. We count them under “dates out of order” and keep them out of the bands: 21 in Colorado, none in Connecticut. A company that traded under another form before it registered, or re-registered later, is aged from the date the registry gives, so its true age may be longer.
What was left out. We left out 1,458 Colorado and 3,163 Connecticut administrative dissolutions. In these, the state dissolved the company for a missed periodic or annual report, and the company may still be trading. They are not in any table.
Size is not screened. Neither register has an employee or size field. The table cannot say which of these companies had 25 or more staff, and we did not try to guess. micro1’s referral page lists 30 or more employees as its company requirement. We refer a closed company to micro1 when it had 25 or more people at its peak. Many of the companies counted here will be below both lines.
Entity type. Grouped from each registry’s codes; every Colorado code in the table is a domestic entity. Other forms are the smaller entity types each registry lists, such as cooperatives and partnership associations in Colorado and general partnerships in Connecticut. The entity rows add up to the year totals: 88,564 in Colorado and 21,775 in Connecticut.
The sources. Both registers are published as open files under a public domain licence: Colorado’s by the Department of State on the Colorado Information Marketplace, Connecticut’s by the Secretary of the State on Connecticut Open Data. Both are updated daily. We publish counts only.
Refresh. The tables are rebuilt monthly from a new read of the database. The next update is due 23 November 2026.
TrustPast
Apply & License Your Data*Purchases are made by micro1 under its data partnership programme. TrustPast is an independent referral partner of micro1.
Sources
- Colorado Information Marketplace, Business Entity Transaction History (Colorado Department of State, public domain)
- Colorado Information Marketplace, Business Entities in Colorado: entity type and formation date (Colorado Department of State, public domain)
- Connecticut Open Data, Connecticut Business Registry - Business Master: registration and dissolution dates (Secretary of the State, public domain)
- micro1, refer a company: company requirements, 30+ employees (micro1.ai)
- micro1, data partnerships: the data it looks for, ownership and PII removal (micro1.ai)
- micro1, how much is your data worth: the years question, anonymisation and redaction agreed before participation (micro1.ai)