CoffeariumBeta
  1. Plant
  2. Farm
  3. Process
  4. Trade
  5. Roast
  6. Brew
  7. Cup
Trade · Supply

Crop year and coffee year

The time frames behind coffee statistics – the ICO coffee year from 1 October, national crop years, USDA marketing years and FAO calendar years – and why one harvest can carry different year labels.

Source-checkedby Coffearium verification pass· Last source-checked 24 Oct 2026· Prepared with AI assistance· 14 sources (13 tier A)· 1 contested· Updated 24 Oct 2026· Literature last searched 24 Oct 2026
Reading depth
Contents
  • The International Coffee Organization (ICO) coffee year runs from 1 October to 30 September; the ICO uses it for the world supply and demand balance and for prices.1,2
  • A crop year is the 12 months from one harvest to the next; the ICO uses crop years that start on 1 April, 1 July or 1 October.2
  • The US Department of Agriculture (USDA) reports each producing country in its local marketing year: July–June for Brazil, April–March for Indonesia, October–September for Colombia.5
  • FAOSTAT, the database of the Food and Agriculture Organization (FAO), reports crop production by calendar year.11
  • Brazil’s crop picked from April to September 2025 is USDA marketing year 2025/26, and the ICO splits each April–March crop year across two coffee years.2,6

Three ways to count a year#

Coffee is picked in different months in each origin, so no single calendar fits every producer. Coffee statistics therefore use several kinds of year. Each dataset names its year in its own way, and the labels look alike.

  • Coffee year. The International Coffee Agreement 2007 defines the coffee year as the period from 1 October to 30 September.1 The International Coffee Organization (ICO) uses it for the world supply and demand balance.2 Coffee year 2022/23, for example, began on 1 October 2022 and ended on 30 September 2023.2
  • Crop year. The ICO defines a crop year as the 12 months from one harvest to the next.2 The crop years it uses start on 1 April, 1 July or 1 October.2
  • Marketing year. The US Department of Agriculture (USDA) reports each producing country in its local marketing year. It starts in October, April or July, depending on the country.5 Countries that do not produce coffee are on an October–September year.5
  • Calendar year. FAOSTAT, the statistics database of the Food and Agriculture Organization (FAO), reports crop production by calendar year.11

A year label such as “2025/26” therefore means different months in different tables. The table below sums up the frames.

Frame Who uses it Months Source
Coffee year ICO: supply and demand balance, exports, prices 1 October–30 September 1,2
Crop year ICO: production of each exporting country From 1 April, 1 July or 1 October 2
Local marketing year USDA: production and trade of each producing country April–March, July–June or October–September 5
Marketing year, non-producers USDA: importing countries October–September 5
Calendar year FAOSTAT January–December 11

Crop years and marketing years by country#

A country’s year usually starts near the beginning of its harvest. Brazil picks robusta from April and arabica from May, and the harvest ends by about September.6 The ICO counts Brazil’s crop year from 1 April.2 The USDA uses a July–June marketing year for Brazil.5,6

Peru harvests from April to September.10 The USDA uses an April–March marketing year for Peru10, and an ICO market report states that Peru’s crop year begins in April.4 Colombia’s main harvest runs from October to December, and its USDA marketing year from October to September.7 Vietnam and Ethiopia are also on October–September marketing years in USDA reports.8,9 Harvest months for each origin are listed in Harvest calendars.

20252026JFMAMJJASONDJFMAMJJASONDCalendar yearFAOSTAT 11,12 20252026ICO coffee year1 Oct–30 Sep 1 2024/252025/2626/27Brazilsources 2,5,6 Perusources 4,10 Colombiasources 5,7 Vietnamsources 8 Ethiopiasources 9,14 Harvest months (2025 harvest)USDA marketing year 2025/26ICO crop year 2025/26
Data table
MonthsSource
Calendar year (FAOSTAT)January–December11, 12
ICO coffee year1 October–30 September1, 2
Brazil: harvestApril–September6
Brazil: USDA marketing yearJuly–June5, 6
Brazil: ICO crop yearApril–March2
Peru: harvestApril–September10
Peru: USDA marketing yearApril–March10
Peru: ICO crop yearbegins in April4
Colombia: harvestOctober–December7
Colombia: USDA marketing yearOctober–September5, 7
Vietnam: harvestlate October–January8
Vietnam: USDA marketing yearOctober–September8
Ethiopia: harvestAugust–January14
Ethiopia: USDA marketing yearOctober–September9
The 2025 harvest of five producing countries against the year frames of three datasets, January 2025 to December 2026. Grey: harvest months; accent: USDA marketing year 2025/26; outline: ICO crop year, where an ICO document states it. Harvest months repeat each year; only the 2025 harvest is shown, and for Colombia only the main harvest.1,2,4,5,6,7,8,9,10,11,14

One harvest, several year labels#

Take the coffee Brazil picked from April to September 2025.6 The USDA counts it in marketing year 2025/26, which runs from July 2025 to June 2026.5,6 In the ICO’s system it belongs to Brazil’s crop year 2025/26, from April 2025 to March 2026.2 That crop year overlaps two coffee years: 2024/25, which ends on 30 September 2025, and 2025/26, which starts on 1 October 2025.1,2

The ICO says so itself. For producers whose crop year starts on 1 April, crop-year production falls across two coffee years.2 To build a coffee-year total, the ICO takes a portion of each of two crop years (see the deep dive).2

Labels from national bodies can differ again. A 2025 USDA report quotes Brazil’s National Supply Company (Conab): 55.7 million bags of production “in 2025”.6 The USDA’s own forecast for marketing year 2025/26 is 65 million bags of 60 kg.6 The report says that Conab uses different methods from the USDA’s Brasília office and has historically given lower estimates; it does not discuss the year labels.6

Why it matters for comparisons#

Two figures with the same label can cover different harvests. A USDA “2025/26” figure for Brazil and an ICO coffee year 2025/26 figure do not cover the same months.2,5 A comparison of USDA, ICO and FAOSTAT world totals is given in The coffee belt.

  • Production. Compare figures from one dataset, with the year frame named. Production by country is set out in Coffee production by country.
  • Exports. The ICO reports exports by coffee year in its monthly market reports.2 A 2014 ICO report gave Brazil’s exports counted from April, the start of its crop year.4
  • Prices. The ICO uses the coffee year for prices as well, so supply, demand and prices refer to the same period.2 The weights of its composite indicator price are set per coffee year.13 How prices form is explained in How a coffee price is made.

Reference values: year frames#

Parameter Value Source
ICO coffee year 1 October–30 September (International Coffee Agreement 2007, Article 2(3)) 1
ICO crop year, definition The 12-month period from one harvest to the next 2
ICO crop year start dates 1 April, 1 July, 1 October 2
ICO crop year, Brazil From 1 April (e.g. 1 April 2022–31 March 2023) 2
ICO crop year, Peru Begins in April (ICO, 2014) 4
USDA marketing year, Brazil July–June 5,6
USDA marketing year, Indonesia Begins in April 5
USDA marketing year, Peru April–March 10
USDA marketing year, Colombia October–September 5,7
USDA marketing year, Vietnam October–September 8
USDA marketing year, Ethiopia October–September 9
USDA marketing year, non-producing countries October–September 5
FAOSTAT crop production Calendar year; series 1961–2024 11

How the ICO converts crop years to coffee years#

The ICO calls its coffee year a marketing year: the year commencing on 1 October.2 To keep the world balance consistent, it converts production from a crop-year basis to this basis, depending on each country’s harvest months.2 The ICO gives Brazil as its example. Brazil’s crop year 2022/23 ran from 1 April 2022 to 31 March 2023 and covered the first half of coffee year 2022/23.2 Crop year 2023/24 ran from 1 April 2023 to 31 March 2024, so it overlaps the second half of coffee year 2022/23 (our reading of the dates; the ICO note says “latter half of coffee year 2023/24” here, which does not fit its own dates or its next sentence, which allocates part of the 2023/24 crop year to coffee year 2022/23).1,2 Coffee year 2022/23 production therefore takes a portion of the 2022/23 and a portion of the 2023/24 crop year.2

202220232024AMJJASONDJFMAMJJASONDJFMCrop year 2022/231 Apr 2022–31 Mar 2023 2 Crop year 2023/241 Apr 2023–31 Mar 2024 2 Coffee year 2022/231 Oct 2022–30 Sep 2023 1,2 a portion of each crop year 2 months inside coffee year 2022/23months in the coffee year before or after
The ICO’s Brazil example: two April–March crop years each overlap coffee year 2022/23, and a portion of each is allocated to it. Schematic, not to scale; the portions follow harvest months, not the number of months drawn.1,2,3

The tool for this is the crop percentage distribution. The ICO’s Statistics Committee describes it as the availability of the crop in each quarter, used to convert supply data from a crop-year basis to a coffee-year or calendar-year basis.3 The resulting figures serve a consistent aggregate balance for analysis. The ICO states that they do not represent the production occurring on the ground within individual countries.2

A hypothetical example shows the arithmetic (our illustration, not ICO data). A country with an April–March crop year harvests 50 million bags in crop year 2025/26. If 40% of that crop is available from October 2025 to March 2026, 20 million bags count in coffee year 2025/26 and 30 million in coffee year 2024/25. Coffee year 2025/26 then also receives a share of the next crop year, 2026/27.

USDA marketing years and FAOSTAT calendar years#

The USDA circular sorts producing countries into three marketing years, beginning in October (Colombia), April (Indonesia) or July (Brazil); non-producing countries use October.5 Country reports state the year in their production figures, for example “MY 2025/2026 (April–March)” for Peru and “MY 2022/23 (Oct-Sep)” for Ethiopia.10,9 The circular does not describe a conversion of these local years to a common year; we found none in the document.5 A USDA world total for “2025/26” is therefore read here as a sum of local marketing years (our reading).

FAOSTAT’s crop production dataset is described with the temporal keyword “calendar year”.11 According to World Bank notes on FAO cereal data, the FAO allocates production to the calendar year in which the bulk of the harvest took place.12 No FAO document consulted for this entry states this rule for coffee. For a harvest that spans the turn of the year, such as Vietnam’s from late October to January8, the months alone do not show which calendar year receives the crop.

Comparing figures across datasets#

The differences in year frame suggest a few checks before two figures are compared (our guidance, drawn from the definitions above).

  1. Name the dataset and its year frame: ICO coffee year, ICO crop year, USDA marketing year or FAOSTAT calendar year.1,2,5,11
  2. Find the country’s start month in that dataset; for Brazil it is April in ICO crop years and July in USDA marketing years.2,5
  3. Check whether the figure is converted. ICO coffee-year production is an analytical conversion, not a count of one harvest.2
  4. Compare exports and prices in the same frame. The ICO reports both by coffee year.2
  5. Keep methods apart from timing. Differences between Conab and USDA estimates for Brazil are put down to method by the USDA.6

Open questions#

  • ICO crop year of each country. The ICO documents consulted for this entry name the three crop-year start dates and give Brazil and Peru as April countries.2,4 None of them lists the crop year of every exporting country.
  • Crop percentage distribution values. The ICO defines the concept3, but no document consulted here publishes the quarterly shares by country.
  • FAOSTAT and coffee harvests across the new year. The FAO rule we found is stated for cereals.12

Key studies#

ICO, 2026 – Coffee Market Report, note to Table 3. Explains why the ICO reports the world supply and demand balance and prices by coffee year from 1 October. Describes its three crop-year start dates and how a Brazilian April–March crop year is split across two coffee years.

International Coffee Organization, February 2026 · ico.org

USDA FAS, 2026 – Coffee: World Markets and Trade. The USDA’s world coffee circular. It lists producing countries by local marketing year, beginning in April, July or October, and puts non-producing countries on an October–September year.

USDA Foreign Agricultural Service, July 2026 · apps.fas.usda.gov

ICO Statistics Committee, 2015 – Data concepts and variables. Defines the variables in ICO statistics, among them the crop percentage distribution: the share of the crop available in each quarter, used to convert crop-year data to coffee or calendar years.

ICO, SC 59/15 · ico.org

Contested
  • Start of Brazil’s year: the ICO counts Brazil’s crop year from 1 April2, while the USDA uses a July–June marketing year for Brazil.5,6 Both frames are definitions, not measurements; neither source says why it chose its start month.
Share or cite this entry

Change log

  • – Removed the figure of 50.1 million bags for Brazil’s exports in coffee year 2024/25; the exports point now only says that the ICO reports exports by coffee year. Reason: Verification found that the ICO February 2026 report gives 50.1 million bags as a 35.1% rise in coffee year 2024/25 in a paragraph that also expects Brazil’s exports to fall once Asian robusta supply recovered in 2024/25; the year label is probably a slip in the report, so the number was not kept.

Sources

  1. AInternational Coffee Organization (2007). International Coffee Agreement 2007 (WP WGFA 3/07 Rev. 8). International Coffee Organization (ICO).
  2. AInternational Coffee Organization (2026). Coffee Market Report – February 2026. International Coffee Organization (ICO).
  3. AInternational Coffee Organization, Statistics Committee (2015). Data concepts and variables used in the statistics of the Organization (SC 59/15). International Coffee Organization (ICO).
  4. AInternational Coffee Organization (2014). Coffee Market Report – December 2014. International Coffee Organization (ICO).
  5. AUSDA Foreign Agricultural Service (2026). Coffee: World Markets and Trade (July 2026). United States Department of Agriculture, Foreign Agricultural Service.
  6. A(2025). Brazil: Coffee Annual (BR2025-0013). USDA Foreign Agricultural Service, GAIN report.
  7. A(2025). Colombia: Coffee Annual (CO2025-0014). USDA Foreign Agricultural Service, GAIN report.
  8. A(2017). Vietnam: Coffee Annual (VM7024). USDA Foreign Agricultural Service, GAIN report.
  9. A(2023). Ethiopia: Coffee Annual (ET2023-0014). USDA Foreign Agricultural Service, GAIN report.
  10. A(2025). Peru: Coffee Annual (PE2025-0018). USDA Foreign Agricultural Service, GAIN report.
  11. AFood and Agriculture Organization of the United Nations (2025). Crop production, yield, harvested area (Global, National – Annual) – FAOSTAT. FAO Data Catalog.
  12. BWorld Bank (data from FAO) World Development Indicators: Cereal yield (kg per hectare), AG.YLD.CREL.KG, indicator notes. World Bank, read through the Datahub copy of the indicator metadata.
  13. A(2026). ICO Indicator Prices – September 2026 (I-CIP). International Coffee Organization.
  14. AMutua, J. (2000). Post Harvest Handling and Processing of Coffee in African Countries, Annex 4: Africa Coffee Producing Countries Profiles. FAO (AGSI).

Tier A · authoritative · B · trade and expert · C · general. Supporting quotes for every source are in the data version of this entry. If a publisher’s site refuses access, try the doi link or “other copies”.

The literature for this entry was last searched on 24 Oct 2026. A statement that no source consulted for this entry shows something refers to the sources above; newer or missing work may exist.

Suggest a change

On Crop year and coffee year. Tell us what you would change and why. A source that supports it helps most.