Estonian CIT from 2027: mid-year entry is being removed, so the decision has a 1 February deadline

The Side Door Is Closing

For four years Estonian CIT has had a forgiving calendar. A company could decide in March, close its books at the end of the month, file ZAW-RD in April and be on the regime from 1 April. Offices relied on it. A client who missed January could still be moved later in the year, at the cost of one extra set of financial statements.

That ends in 2027. The amendment to the PIT and CIT acts (the government draft known as UD116, now Sejm print 3137) removes mid-year entry. From 2027 a company can choose Estonian CIT only from the start of its tax year. For a calendar-year company that means ZAW-RD filed with the tax office by the end of January 2027. 31 January 2027 is a Sunday, so the real deadline is Monday, 1 February 2027.

Miss it, and the next chance is 1 January 2028. A decision that used to have twelve entry points a year now has one.

A note on status. As I write this (24 October 2026) the bill is still in Parliament, with entry into force planned for 1 January 2027. Until it's published in Dziennik Ustaw, treat the mid-year rule as planned, not final. For planning, it makes no difference: if your clients are going to change regime in 2027, plan for 1 January.

What 1 February Is Worth: One Company, Two Regimes

A small-taxpayer company, sp. z o.o., two shareholders who are individuals, three full-time employees, 400 000 zł of profit a year, all of it paid out. I'll run the standard comparison you'll find in most write-ups: tax only, with 400 000 zł as the base in both regimes, even though they measure profit slightly differently.

Standard CIT at 9%. 36 000 zł of CIT, 364 000 zł of dividend, 19% PIT on that is 69 160 zł. Total: 105 160 zł, an effective 26.29%.

Estonian CIT at 10%. 40 000 zł of CIT on the distributed profit. The shareholders pay 19% PIT, 76 000 zł, minus 90% of the company's CIT (36 000 zł), so 40 000 zł. Total: 80 000 zł, an effective 20%.

Difference: 25 160 zł a year. For a company that isn't a small taxpayer the rates are 20% Estonian against 19% standard, with a 70% credit, which gives 25% against 34.39%: about 37 600 zł a year on the same profit.

Now take the same company, but say it reinvests everything and pays nothing out in 2027. Standard CIT takes 36 000 zł in 2027 regardless. Estonian CIT takes nothing until the profit leaves the company. For a client buying machines or building a warehouse, that deferral is often the real reason to switch, more than the lower combined rate.

Miss 1 February, and a company like this pays standard CIT for the whole of 2027. That's 25 000 zł if it distributes, 36 000 zł of cash tied up if it reinvests.

Who Can't Use It, or Shouldn't

The rate comparison is the easy half. The office's real work is the conditions, and most of them sit in data the office already holds:

• Shareholders. Only individuals (the family-foundation case is a narrow exception). A company with another company on its shareholder list is out.
• Holdings. The company itself can't hold shares in other companies or units in investment funds.
• Employment. In most cases at least three full-time employees who aren't shareholders, or the equivalent salary spending under other contracts. The amendment loosens this: employment contracts and other contracts can be combined to meet the threshold. Newly formed companies get relief for the first years.
• Passive revenue. Interest, royalties, related-party transactions and similar must not exceed 50% of revenue.
• Accounts. Full books under the Accounting Act, and an approved financial statement. For entry on 1 January, the statement for 2026 does that job, so no extra closing is needed.

And one condition that isn't a condition but often decides the answer:

• Unused losses. Losses from before the regime can't be deducted under Estonian CIT, and they don't come back if the company leaves. Take a company with 120 000 zł of unused loss from 2024: on standard CIT, 2027 might cost almost nothing. Put it on the Estonian regime and that loss is gone.

Then there's the deferred tax picture and the transition adjustments, which matter more for companies with large provisions or fixed-asset differences. That's a conversation for the tax adviser, not a checklist item. But the office should know which clients have them before the conversation starts.

Small Taxpayer Status for 2027

The 10% rate depends on small-taxpayer status, which for 2027 is set by 2026 revenue: up to the equivalent of 2 million euro, including VAT, at the NBP rate from the first working day of October. That's 8 754 000 zł for 2027, as I worked out in the 2027 limits post. A company growing past that line in 2026 compares 20% against 19%, not 10% against 9%, and the numbers come out differently.

How Many Clients This Is

A typical office I work with has 20 to 30 companies on full books. Usually a handful are already on the Estonian regime, and the question is the rest. In my experience the shareholder condition alone removes a third of them, and the employment condition removes another third. That leaves five to eight companies where the comparison is worth running properly.

Running it by hand means four sources per client: KRS for the shareholders, payroll for the headcount, the general ledger for the passive revenue and the CIT-8 archive for the losses. Then the profit-distribution plan, which lives in the owner's head. An hour per company if everything is where it should be. In November, alongside the form-of-taxation conversations with sole traders, nobody has that hour.

The usual mistakes:

• The decision gets pushed to "after the year-end". Under the old rules that was annoying but recoverable. Under the new ones it means waiting until 2028.
• The comparison ignores the loss. The 20% looks better than 26.29% until a 120 000 zł loss enters the calculation.
• Nobody checks the shareholder list. A holding company added as a shareholder two years ago closes the regime, and it's easy to miss.
• Dividend plans aren't asked about. A company that reinvests everything and a company that pays everything out get different answers, and the office often doesn't know which one it's talking to.

What the Agent Prepares

An agent with read access to the office's data can assemble the list in November, as part of the month-end close the office already runs:

• Reads the shareholder list from KRS for each company client and flags anything that isn't an individual.
• Reads headcount and payroll for 2026 and checks the employment threshold month by month, not only at year-end.
• Computes passive revenue from the general-ledger accounts the office maps once (interest, royalties, related-party revenue) as a share of total revenue.
• Reads unused losses from past CIT-8 returns and the 2026 revenue for small-taxpayer status.
• Runs the comparison for both regimes in two variants, full payout and full reinvestment, and writes one line per company: "meets all conditions, about 25 000 zł a year at full payout, ZAW-RD by 1 February" or "has 120 000 zł unused loss, standard CIT cheaper for 2027".

The agent doesn't choose the regime and doesn't file ZAW-RD. The accountant and the tax adviser decide with the owner. It has to be a person: half the answer is what the owner plans to do with the money.

One Thing to Do This Week

List your full-books companies. Next to each, write two words: "individuals" or "not" for the shareholders, and "yes" or "no" for three non-shareholder employees through 2026. Every company with "individuals" and "yes" should hear from you in November, not in January.

If you'd like to see this list built from your own clients' KRS data, payroll and ledgers, write to me. I'll take one month of real data and show you which companies the agent flags and why.