Why Cambodia’s Coffee Production Numbers Don’t Match — And What That Reveals About an Emerging Origin

How much coffee does Cambodia actually produce?

Depending on which public report you open, the answer can look like 300–400 tonnes, 675 tonnes, 725 tonnes, 1,333 tonnes, more than 1,600 tonnes, or roughly 10 percent of a 20,000-tonne domestic market.

Those figures cannot all be treated as the same measurement.

That does not automatically mean someone is lying or that Cambodia has no useful coffee data. It means the public record mixes different years, different geographies, different stages of farm development and, in some cases, poorly defined units such as “coffee beans” without clearly stating whether the figure refers to fresh cherry, dried coffee, parchment or export-ready green coffee.

For an emerging origin, this is more than a statistical nuisance.

Production data affects investment decisions, farm expansion, processing capacity, import-replacement policy and the credibility of origin claims. If the numbers are not defined carefully, a buyer can mistake planted hectares for productive hectares, a journalist can compare a provincial estimate with a national figure, and a market report can turn a rough demand ratio into a precise-sounding production number.

OCC’s approach is therefore not to choose the largest or smallest figure and declare it “correct.”

The better method is to build a source ledger: year, reporting date, geography, area definition, production unit, source and unresolved questions.

The short answer: the latest ministry-linked figure is 1,333 tonnes, but it does not erase earlier data

The most recent production figure we located that is directly attributed to Cambodia’s Ministry of Agriculture, Forestry and Fisheries comes from a Phnom Penh Post report published on 6 June 2026.

Ministry spokesperson Khim Finan said coffee cultivation had expanded to 705 hectares nationwide, with 563 hectares under harvest, producing 1,333 tonnes of beans at an average reported yield of 2.36 tonnes per hectare.

For a current country-level reference, that is the strongest number in the public material reviewed for this article because the geography is national, planted and harvested area are separated, and the production figure is tied to a named ministry spokesperson.

Even here, one important definition remains unclear in the published report: “beans” is not enough to identify the exact physical form of coffee represented by the 1,333 tonnes.

That matters because coffee changes mass dramatically between fresh fruit and a prepared green-bean product. Without a defined product stage, production numbers should not automatically be used for green-coffee supply calculations.

A source-by-source comparison

The table immediately shows why a single timeless sentence such as “Cambodia produces X tonnes of coffee per year” is risky.

The figures are not observations of the same thing at the same moment.

The 2023 Mondulkiri numbers are less contradictory than they first appear

The August 2024 Phnom Penh Post report is a good example of how reading the denominator changes the interpretation.

Mondulkiri provincial agriculture director Song Kheang reported that coffee cultivation covered 525 hectares in 2023 and involved 365 farming families. The same report said that only 346 hectares were currently being harvested and that average yield was 2.096 tonnes per hectare. Total harvest was reported at about 725 tonnes per year.

Those numbers fit together closely.

The productive 346 hectares multiplied by the reported 2.096 tonnes per hectare yields approximately the reported 725 tonnes.

So the relevant yield denominator is not the entire 525 hectares planted. A substantial part of the area was still immature or otherwise not yet under harvest.

If someone takes “525 hectares” from one sentence and “725 tonnes” from another, then calculates a province-wide yield of roughly 1.38 tonnes per hectare, they have created a new statistic that the source itself did not report.

This distinction matters especially for coffee because new plantations take time before entering meaningful commercial production.

When cultivation expands quickly, total planted area can rise well before production catches up.

National and provincial figures are not interchangeable

Another recurring mistake is comparing Mondulkiri with Cambodia as if the geographic labels were identical.

The 2021 figure of 675 tonnes cited from Ministry of Agriculture data is national.

The 2023 figure of approximately 725 tonnes is Mondulkiri-specific.

At first glance it can seem impossible for one province in 2023 to produce more than the whole country did in 2021. But there is no mathematical contradiction once the two-year gap and sector expansion are acknowledged.

Production can increase materially between harvests, especially from a small base and when previously planted coffee begins bearing.

What would be a mistake is presenting the 675 and 725 figures as competing estimates for the same year.

They are not.

The 2020–2021 decline is real within that reporting series

Multiple reports citing the Ministry of Agriculture stated that Cambodia produced 940 tonnes of coffee in 2020 and 675 tonnes in 2021, a decline of more than 28 percent.

Those numbers are useful because they belong to the same national reporting sequence and were presented as directly comparable.

What they do not tell us is why production fell.

The cited articles do not provide enough detail to confidently attribute the decline to weather, farm abandonment, crop cycles, measurement changes or another cause.

A responsible article should therefore preserve the observation—940 to 675 tonnes—without inventing the explanation.

This is a general rule for Cambodia coffee statistics: missing methodology is not permission to supply a narrative from intuition.

The 300–400 tonne estimate is a warning about source hierarchy

A December 2025 report on a coffee-value-chain forum at the Royal University of Agriculture described the Kingdom’s production as an estimated 300–400 tonnes annually from around 500 hectares, mainly in Mondulkiri.

That is difficult to reconcile with the 2023 Mondulkiri agriculture-department figure of around 725 tonnes or the June 2026 ministry figure of 1,333 tonnes nationally.

There are several possible explanations.

The estimate may refer to a narrower product form. It may use a different crop year. It may count a subset of commercially traded output. It may simply be a rough forum estimate based on a different dataset.

The source available to OCC does not provide enough methodology to determine which explanation is correct.

The right editorial response is therefore not to average 350, 725 and 1,333 into a synthetic “Cambodia production estimate.”

It is to rank the evidence.

A named ministry spokesperson giving national planted area, harvested area, output and yield is more useful for a current production reference than an unattributed broad estimate in a conference report.

That does not make the weaker estimate worthless. It makes its uncertainty visible.

The >1,600-tonne Mondulkiri figure is currently unresolved

A 2026 secondary coffee-industry article says a 25 May Khmer Times report placed Mondulkiri at more than 580 hectares of coffee and more than 1,600 tonnes of annual production, with authorities planning to add another 2,000 hectares over coming years.

The figure has since been repeated by other websites.

OCC was able to locate the secondary reporting and repeated references, but not a directly accessible original article containing enough methodology to resolve the production definition.

More importantly, the figure creates an obvious conflict with the Ministry-linked national data published less than two weeks later: 705 hectares nationwide, 563 hectares under harvest and 1,333 tonnes produced.

Mondulkiri cannot logically produce more than 1,600 tonnes while the entire country produces 1,333 tonnes if both reports refer to the same time period and the same physical coffee product.

At least one of those conditions must differ—or one figure is inaccurate.

Possible differences include reporting period, crop form, forecast versus realised production, geographical scope or source error.

Until the underlying table or official provincial record is available, OCC treats the >1,600-tonne figure as unresolved, not as a replacement for the ministry-linked national number.

That is exactly the kind of uncertainty an evidence-led origin publication should preserve.

“Coffee beans” is not a sufficient unit

Many news articles use the phrase “coffee beans” casually.

For market analysis, that wording is inadequate.

Coffee can be counted as fresh cherry immediately after harvest, dried cherry, parchment coffee, hulled green coffee, roasted coffee or even finished retail product. Each stage has different mass and commercial meaning.

A farmer-purchase story may use a weight from cherries delivered at farm gate. A ministry crop report may use another agricultural convention. A trader may quote green coffee. A consumer-demand estimate may be based on roasted or imported product.

If the reporting does not define the stage, two identical tonne figures may represent very different amounts of export-ready green coffee.

This is also why price data require the same discipline. A price per kilogram for fresh cherry cannot be compared directly with a price per kilogram for prepared green coffee.

The underlying problem is not arithmetic. It is classification.

Planted area and harvested area need separate columns

The 2026 Ministry-linked figure provides a useful template: 705 hectares cultivated, 563 hectares under harvest.

Those are two different statistics and both are necessary.

For an established annual crop, the gap may be small. In an expanding perennial crop such as coffee, it can be substantial.

New coffee trees do not contribute full commercial production immediately. Expansion programmes can therefore make the planted-area line rise before the harvest line follows.

If Cambodia executes plans to expand coffee cultivation materially, this distinction will become even more important.

Future reporting should ideally provide at least:

• total planted area;
• bearing or harvested area;
• newly planted immature area;
• number of farms or households;
• production volume;
• crop form represented by the volume; and
• yield calculated only on the relevant productive area.

Without these definitions, growth can look either much faster or much slower than it really is.

The domestic-demand ratio is also an estimate, not a production census

Several 2026 reports cite annual Cambodian coffee consumption at around 20,000 tonnes and say domestic production meets only about 10 percent of demand.

Taken literally, 10 percent of 20,000 tonnes would imply around 2,000 tonnes of domestic supply.

That is higher than the ministry spokesperson’s 1,333-tonne production figure.

Again, this does not necessarily prove one claim is false.

“About 10 percent” is clearly a rounded market estimate, while “20,000 tonnes” may itself use a different product basis from agricultural production statistics. Imports, carryover inventory, roasted-product equivalence and informal flows can also complicate a direct comparison.

The responsible interpretation is qualitative: Cambodia remains heavily dependent on imported coffee and domestic production covers only a minority of demand.

The public data reviewed here do not justify converting the 10 percent statement into a precise production census.

Why young industries often have messy statistics

Mature commodity sectors tend to have standardised reporting because governments, exporters, exchanges, trade associations and buyers have spent decades agreeing on definitions.

Cambodia’s modern coffee sector is not there yet.

Production is concentrated in a relatively small number of areas, farms are expanding, new private purchasing networks are forming, processing capacity is changing and the policy focus on coffee has intensified only recently.

Different institutions naturally collect different data for different purposes.

A provincial agriculture department may track cultivated land and farmer households.

A company may track contracted suppliers and purchased cherry.

A development organisation may track hectares under sustainable management.

A coffee association may estimate domestic demand.

A ministry may publish national crop production.

A news outlet may compress all of those into the word “production.”

The result is apparent contradiction.

The solution is not necessarily one central number. It is a shared vocabulary.

Better coffee data would create economic value

Statistical quality can sound like an academic concern. For Cambodia it has direct commercial consequences.

A processor deciding whether to build a mill needs to know how much harvest is available within realistic collection distance.

A lender financing plantation expansion needs bearing-area and yield assumptions.

A buyer considering a multi-year sourcing programme needs to understand commercial volume, not just planted hectares.

A government evaluating import substitution needs comparable production and consumption units.

A farmer deciding whether to plant coffee needs credible expectations about the market rather than a headline suggesting supply is either abundant or almost nonexistent.

Better data therefore functions as infrastructure.

It reduces uncertainty in the same way that better roads, laboratories or processing facilities do.

What OCC will use as the current baseline

For current editorial work, OCC will use the June 2026 Ministry-linked figure as the primary national baseline unless a newer, equally or more authoritative dataset becomes available:

Cambodia: 705 hectares under coffee cultivation; 563 hectares under harvest; 1,333 tonnes produced; reported average yield 2.36 tonnes per harvested hectare.

Every reuse of that number should retain its date and source context.

For Mondulkiri historical context, OCC will retain the provincial 2023 figures reported in August 2024:

525 hectares total cultivation; 346 hectares under harvest; approximately 725 tonnes; 365 farming families; reported yield 2.096 tonnes per harvested hectare.

The >1,600-tonne Mondulkiri claim will remain labelled unresolved until the original statistical basis can be verified.

The 2020 and 2021 national ministry figures—940 and 675 tonnes—remain useful historical points, not current output estimates.

A data protocol Cambodia’s coffee sector could adopt

A simple reporting standard would eliminate much of the confusion.

Every public production number should ideally state five things next to it:

1. Crop year. Which harvest does the number describe?

2. Geography. Cambodia, Mondulkiri, a district, a company supplier network or one farm?

3. Area status. Planted hectares or bearing/harvested hectares?

4. Product form. Fresh cherry, dried cherry, parchment, hulled green coffee or roasted coffee?

5. Source and method. Administrative records, farmer survey, company purchases, estimate or forecast?

With those five labels, many apparently conflicting statistics can coexist without confusion.

Without them, even accurate figures can be misleading.

The mismatch itself tells us something about Cambodia

Cambodia’s coffee numbers do not yet behave like the statistics of a mature producing origin.

That is not a reason to dismiss the sector.

It is evidence of where the sector is in its development.

Coffee cultivation is expanding from a small base. Some planted land is still maturing. Public institutions, private companies and development organisations are collecting different pieces of the picture. Market demand is much larger than domestic supply. And international attention to Cambodian Fine Robusta is growing faster than the country’s statistical architecture.

The next stage of origin building therefore requires two kinds of quality control.

One is in the cup.

The other is in the spreadsheet.

For Cambodia to become a credible Fine Robusta origin, both need to improve together.

Sources and further reading

• Phnom Penh Post. “Bean counters: Demand sees surge in coffee production.” 6 June 2026. Ministry spokesperson: 705 ha nationwide; 563 ha harvested; 1,333 t; 2.36 t/ha.
• Phnom Penh Post. “Mondulkiri coffee farmers urged to expand after ‘Fine Robusta’ recognition.” 16 August 2024. 2023 Mondulkiri figures: 525 ha cultivated; 365 farming families; ~725 t; 346 ha under harvest at 2.096 t/ha.
• Phnom Penh Post. “Coffee roasters KOFI sets sights on export markets.” 2022. Cites Ministry figures of 940 t in 2020 and 675 t in 2021.
• Cambodian government/7 January reporting. “Collaboration to Boost Coffee Development in Mondulkiri Province.” 22 February 2023. Repeats Ministry 2020–2021 figures.
• Khmer Times reporting republished by Vietstock. “Cambodia smells opportunity in coffee value chain.” 29 December 2025. Estimate of 300–400 t annually from around 500 ha.
• CoffeeFactz. “Mondulkiri coffee partnerships aim to scale Fine Robusta.” 28 May 2026. Secondary record of a Khmer Times 25 May report citing >580 ha and >1,600 t in Mondulkiri; treated by OCC as unresolved pending primary statistical documentation.

Related OCC reading:

• Robusta Cambodia: Fine Robusta, Mondulkiri, Growing Conditions, and Quality Evidence
• Cambodia Does Not Need to Be the Next Vietnam: A Different Strategy for Fine Robusta
• The Ecosystem Behind Cambodia Fine Robusta
• Mondulkiri Coffee: Cambodia’s Fine Robusta Origin Guide for Buyers