Estonian felling fell to 10.57 million cubic metres in 2025, down 24.3% on the year and the lowest annual figure since 2010 (Statistics Estonia, table MM04). The headline is unusual on its own. What makes it a diligence question rather than a news item is where the drop sits: state forest barely moved, and private forest absorbed effectively all of it.
State forest output fell 8.0%, from 3.83 to 3.52 million cubic metres, which is close to its recent drift. Private forest fell 30.7%, from 10.01 to 6.94 million, in a single year. Estonia did not run out of forest in 2025. Several tens of thousands of private owners decided not to sell.
Data table
| Period | Private forest | State forest |
|---|---|---|
| 2021 | 9,386,931 | 4,130,563 |
| 2022 | 10,080,735 | 4,083,329 |
| 2023 | 8,328,628 | 4,105,815 |
| 2024 | 10,012,347 | 3,825,877 |
| 2025 | 6,939,242 | 3,518,224 |
One year undoes a decade of ramp
The longer series puts the scale in context. Estonian felling roughly doubled between 2010 and its 2018 peak of 15.64 million cubic metres, then drifted in a band above 12 million for six years. The 2025 figure sits below that band entirely, back at a level last seen in 2010.
Data table
| Period | All owners |
|---|---|
| 2010 | 10,471,310 |
| 2011 | 10,769,352 |
| 2012 | 10,770,760 |
| 2013 | 11,195,661 |
| 2014 | 13,158,671 |
| 2015 | 12,598,158 |
| 2016 | 13,819,442 |
| 2017 | 13,867,275 |
| 2018 | 15,641,648 |
| 2019 | 13,018,308 |
| 2020 | 13,203,374 |
| 2021 | 13,606,895 |
| 2022 | 14,301,273 |
| 2023 | 12,549,414 |
| 2024 | 13,959,167 |
| 2025 | 10,566,060 |
A single year is not a trend, and this series has moved sharply before. Felling fell 33% in 2005 and did not regain its 2004 level until 2010, five years later. An abrupt drop in this series is not automatically a short one, which is the reason the relevant question for a buyer is not whether 2025 was unusual, but whether the conditions that produced it persist into the year in which a purchased parcel would actually be harvested.
The ownership mix shifted with it
Because private forest fell so much further than state forest, the composition of Estonian supply moved as well. Private forest supplied 71.7% of national felling in 2024 and 65.7% in 2025, the lowest private share in the five-year window. State forest output is set by a management plan rather than by an owner reading the market, so it acts as a floor under national supply and its share rises automatically whenever private owners step back.
That matters for anyone modelling regional roundwood availability. In a year when private supply contracts, the residual market is more heavily state-supplied, more predictable in volume, and less responsive to price. The stability is real but it belongs to a counterparty that is not selling land.
The cut that stopped is the discretionary one
Breaking private felling by type shows this was a decision, not a constraint. Clear felling, the operation that realises the standing value of a mature stand, fell 31.2%, from 8,255,000 to 5,678,406 cubic metres. That single category accounts for 76% of the entire national contraction.
Sanitary felling moved least. It fell 18.7%, from 229,069 to 186,194 cubic metres. Sanitary felling is the least discretionary category on the list: it removes damaged, diseased or windthrown timber, and an owner who has it generally has to do it. The pattern across the two is coherent. Where owners had a choice, they postponed. Where they did not, they cut.
Every county fell, but not evenly
All fifteen counties recorded lower private felling. The spread is wide enough to matter for anyone underwriting a specific parcel rather than a national average: Hiiu county lost 60.6% of its private harvest, while Jõgeva lost 14.6%. The national figure of 30.7% sits between them and describes neither.
Data table
| Area | 2024, m³ | 2025, m³ | Change |
|---|---|---|---|
| Hiiu | 385,268 | 151,945 | -60.6% |
| Põlva | 531,743 | 321,497 | -39.5% |
| Lääne | 408,566 | 248,364 | -39.2% |
| Harju | 737,070 | 490,077 | -33.5% |
| Ida-Viru | 378,521 | 253,220 | -33.1% |
| Viljandi | 993,581 | 667,382 | -32.8% |
| Tartu | 623,143 | 422,941 | -32.1% |
| Lääne-Viru | 632,797 | 439,947 | -30.5% |
| Võru | 973,067 | 681,242 | -30.0% |
| Pärnu | 1,138,085 | 803,960 | -29.4% |
| Järva | 730,927 | 516,472 | -29.3% |
| Saare | 760,633 | 571,323 | -24.9% |
| Rapla | 632,864 | 477,562 | -24.5% |
| Valga | 510,528 | 401,964 | -21.3% |
| Jõgeva | 575,554 | 491,346 | -14.6% |
A county-level read is the beginning of a question, not an answer. Counties differ in species mix, ownership fragmentation, road access and the age structure of their stands, and any of those can produce a deeper or shallower response to the same market signal. The number tells you where to look.
It is not a damage story
One explanation can be dismissed with the same source. If forests had been hit hard, sanitary felling would have risen and standing volume would have been destroyed rather than withheld. The opposite happened. Destroyed forest stands totalled 1,758 hectares in 2025, down from 3,268 hectares at the 2023 peak, a fall of 46.2% (table KK513). Physical risk eased over the same period in which harvesting collapsed.
So the timber is still standing. For a buyer, that is the material point: the 2025 contraction did not remove inventory from the Estonian resource, it deferred its conversion into cash.
The forward signal is weaker than the headline
Reforestation gives the counterweight. Replanted and regenerated area came to 8,815 hectares in 2025, down from 10,462 hectares in 2022 and the lowest in the 8-year window (table MM10). Reforestation follows felling with a lag, so some of that decline is simply less ground to replant. The part that is not mechanical matters more, because establishment decisions taken now set the age structure that a buyer inherits for the next rotation.
What the statistics do not say
Statistics Estonia publishes felling volumes, not roundwood prices, and the felling series is compiled from felling documentation rather than mill intake. Three limits follow, and they should be stated before anyone builds a thesis on the number.
First, the cause is outside this dataset. A deferral of discretionary harvest is consistent with weak roundwood prices, with rising harvesting and haulage costs, with owners waiting out a soft market, and with regulatory or certification friction. The MM04 series cannot separate them, and neither can any other table in the statistical database. Establishing which mechanism dominates needs price and cost evidence from another source.
Second, documentation-based volumes and inventory-based estimates are different measurements. Statistics Estonia publishes a parallel National Forest Inventory series (table MM03) built on statistical sampling rather than paperwork. The two do not have to agree, and a divergence between them is itself informative.
Third, recent years revise. Annual forestry figures are restated as documentation completes, so the 2025 value should be treated as provisional until at least one further release has passed.
Frequently asked questions
How much timber was felled in Estonia in 2025?
10,566,060 cubic metres of gross felling outturn across all owners and all felling types (Statistics Estonia, table MM04). That is 24.3% below 2024 and the lowest annual total since 2010. The figure is compiled from felling documentation and is provisional until later releases confirm it.
Why did Estonian felling volumes fall in 2025?
The statistics show where the fall happened but not why. Private forest fell 30.7% while state forest fell 8.0%, and within private forest the decline was concentrated in clear felling, the most discretionary operation, which fell 31.2%. Sanitary felling, the least discretionary, fell only 18.7%. That is consistent with owners postponing harvest rather than being unable to harvest, but Statistics Estonia does not publish roundwood prices, so separating weak prices from rising costs needs evidence from another source.
Which Estonian county saw the largest fall in private forest felling?
Hiiu county, where private forest felling fell 60.6%, from 385,268 to 151,945 cubic metres between 2024 and 2025. Jogeva county fell least at 14.6%. All fifteen counties declined.
Did storm or pest damage cause the 2025 drop?
No. Destroyed forest stands totalled 1,758 hectares in 2025, down 46.2% from the 3,268 hectares recorded at the 2023 peak (table KK513). Physical damage fell over the same period in which harvesting contracted, and sanitary felling declined rather than rose. The standing timber was withheld, not lost.
What does the 2025 felling drop mean for buying Estonian forest land?
A growth-and-yield discounted cash flow is sensitive to the assumed date of first commercial harvest, and in a market where owners across every county have just deferred that decision, the assumed harvest date should be tested rather than inherited. Transaction comparables drawn from 2025 also come from a thinner and more selective pool.
Why this matters before a purchase
A withheld harvest is not a valuation event by itself. It becomes one through timing. A growth-and-yield discounted cash flow is sensitive to when the first commercial cut is assumed to happen, and a market in which owners across every county have just postponed that decision is a market where the assumed harvest date deserves testing rather than inheriting. Two parcels with identical standing volume can carry materially different values if one assumes conversion this year and the other in four.
The second implication is about comparables. Transaction evidence drawn from a year in which discretionary selling fell by roughly a third is drawn from a thinner and more selective pool. Owners who transacted in 2025 were, almost by definition, less able to wait than those who did not. That is a selection effect, and it argues for widening the comparison window rather than treating recent prints as representative.
None of this changes the structural case for Estonian forest land. It changes the questions worth asking about a specific parcel: how mature is the standing stock, what does the harvest schedule assume, and does the comparable set span more than one point in the cycle. Those are parcel-level questions, and they are what the methodology is built to answer.
Figures in this article are drawn from the Statistics Estonia database on 25 August 2026 and carry the table identifiers used to retrieve them, so any of them can be re-queried at source. Data is reused under CC BY-SA 4.0.