CoffeariumBeta
  1. Plant
  2. Farm
  3. Process
  4. Trade
  5. Roast
  6. Brew
  7. Cup
Post-harvest processing · Harvest · Step 1 of process

Harvesting and cherry sorting

How coffee cherries are picked – by hand, by stripping or by machine – how ripeness is judged, how fast the fruit must reach the mill, and how cherries are sorted before processing.

Source-checkedby Coffearium verification pass· Last source-checked 8 Oct 2026· Prepared with AI assistance· 13 sources (12 tier A)· 3 contested· Updated 8 Oct 2026
Reading depth
Contents
  • The Codex code of practice distinguishes four harvesting systems – single-pass stripping, multi-pass stripping, multi-pass selective picking and mechanical harvesting.1
  • In a Cenicafé study, semi-ripe to overripe fruit gave the best cups and green and dry fruit poor ones; Cenicafé guidance says quality deteriorates from 2.5% green fruit in the harvest.3,7
  • Cherries should be processed as soon as possible – the same day, within 10 hours in a Kenyan example, or within 6 hours in Colombian guidance. Fruit kept in bags or heaps heats and ferments.1,7,12,13
  • Before processing, unripe, overripe, dried and diseased cherries, leaves and twigs are removed by hand and in water, where floaters are skimmed off.1,2,11

How ripeness is judged#

A coffee cherry is ready when its skin has turned from green to its ripe colour: bright red to deep red, or yellow.2 The colour change comes from the replacement of chlorophyll in the skin by red flavonoid pigments.4 As the fruit ripens it also becomes softer and sweeter: in a Cenicafé study, firmness fell steadily and soluble solids (°Brix) rose through ripening, with the highest mean value in overripe fruit.3

Ripeness is classified by colour, in research with a colorimeter.5 Softness and °Brix are also measured in research;3 the sources opened for this entry give no agreed °Brix target for picking (see Deep dive).

123456
  1. Green, 182–203 days. Three green stages and a green-yellow one; fresh weight 1.33 g per fruit in the early green stages. Green fruit gave low cup scores.3
  2. Semi-ripe (pintón), 210 days. The skin turns as red flavonoid pigments replace chlorophyll.3,4
  3. Ripe, 217 days. Fresh weight 1.99 g per fruit.3
  4. Overripe, 224 days. Fresh weight 1.88 g; soluble solids (°Brix) at their highest mean value.3
  5. Dry, 231 days. Fresh weight 0.66 g. Dry fruit gave low cup scores.3
  6. Best cups. Semi-ripe to overripe fruit (210–224 days) gave the best cups. Over these stages firmness fell from 23.63 to 8.78 N along the polar axis.3
Ripening stages of Coffea arabica var. Colombia, sampled weekly from 182 to 231 days after flowering in a Cenicafé study.3 Schematic, not to scale.

Harvest methods#

The Codex code of practice names four systems:1

  1. Single-pass stripping – all fruit-bearing branches are stripped at once.1
  2. Multi-pass stripping – only branches bearing mainly ripe cherries are stripped on each pass.1
  3. Multi-pass selective picking (finger picking) – only ripe cherries are picked.1 It needs several picking rounds and a lot of labour, but gives uniform cherries.2 Kenya, for example, harvests by selective picking of ripe berries.11
  4. Mechanical harvesting – machines take the fruit all at once.1

Stripping, by hand or by machine, mixes unripe and overripe cherries into the harvest, and the lot may miss the desired quality.2 In a Brazilian trial on microterraces in sloping land, one unilateral harvester did the work of about 24 manual pickers.10

Fruit on the ground. The Codex code advises against collecting fallen cherries, especially in humid conditions, because fungal growth on them can lead to ochratoxin A. In wet climates only same-day collection from the ground is acceptable.1

From tree to mill#

Coffee cherries should be processed as soon as possible after harvesting.1 Guidance agrees on the direction but differs in detail:

  • Codex: as soon as possible; holding fruit in bags or heaps, sometimes for up to a week, causes high temperatures and quick fermentation, quality loss and a higher risk of ochratoxin A.1
  • FAO manual for Lao PDR: process the same day, and do not mix with the previous day’s harvest.12
  • Kenya: cherry is taken to the factory immediately after harvesting,11 and pulping must be done within 10 hours of harvesting.13
  • Cenicafé (Colombia): do not leave harvested cherry more than 6 hours before pulping.7 Leaving cherry more than 12 hours or overnight is listed as a quality risk.8

Sorting before processing#

Low-quality cherries – unripe, overripe or diseased – should be removed before processing.1 Two steps are common:

  • Hand sorting. In Kenya the cherry is sorted before pulping to remove immature, diseased, insect-damaged and dry berries, and leaves, twigs and other foreign matter.11
  • Flotation. Cherries are put in water and the floaters are separated: overripe and undeveloped cherries, sticks and leaves.2 Codex names visual sorting or water separation as the two ways to remove low-quality cherries.1 Not separating floaters, borer-damaged fruit, immature fruit and debris is listed as a quality risk.8

For natural coffee the same rule applies before drying: immature and over-mature cherries should be sorted out.1

Ripening physiology#

Timing. In the Cenicafé study on var. Colombia, fruit was sampled weekly from 182 to 231 days after flowering. Green stages ran from 182 to 203 days, semi-ripe (pintón) came at 210 days, ripe at 217 days, overripe at 224 days and dry at 231 days.3 A later Cenicafé paper cites maximum fruit development between 220 and 240 days after anthesis.4

Mass and firmness. Fresh weight rose from 1.33 g in the early green stages to 1.99 g at ripeness, fell to 1.88 g when overripe and to 0.66 g in dry fruit.3 Between the semi-ripe and overripe stages (210–224 days), firmness ranged from 23.63 to 8.78 N along the polar axis and from 23.32 to 6.93 N across the equator, falling as the fruit ripened.3

Pigments. Red colour comes from flavonoid pigments that replace chlorophyll in the skin.4 Anthocyanins kept rising after ripeness: from 1.52 µg per g fresh weight in semi-ripe fruit and 8.26 µg/g in ripe fruit to 47.15 µg/g in overripe fruit.3

Sugars. Sugar content rises in both pericarp and seeds during ripening, and sucrose can make up 73% of the total sugars in the seeds.4 But the fruit and the seed behave differently late in ripening. In a 2024 study of six red-colour classes, sucrose nearly doubled from the lightest to the darkest class in the skin-and-pulp and in the juice, while it did not change significantly in the beans.5 Total soluble sugars were highest in the outer fruit and lowest in the bean.5 Across three mature classes (red-orange, crimson, red-wine), a 2023 study found no differences in organic acids, lipids, chlorogenic acids, proteins, alkaloids or sucrose in the green coffee; only fructose and glucose rose.4 The practical point: more sugar in the pulp does not mean more sugar in the seed.5

Ripeness and cup quality: measured effects#

Across the whole ripening range. Green fruit (182–203 days after flowering) and dry fruit (231 days) gave low cup scores; semi-ripe to overripe fruit (210–224 days) gave the best.3

Within mature fruit. Red-orange, crimson and red-wine fruit of var. Castillo scored 82.73, 82.79 and 82.43 SCA points, with no significant difference in any sensory attribute.4 Darker fruit had less mucilage (10.07% against 15.44%).4

8081828384SCA pointsAxis starts at 80Red-orange82.734 Crimson82.794 Red-wine82.434
Data table
SCA pointsSource
Red-orange82.734
Crimson82.794
Red-wine82.434
Cup scores (SCA) of three mature colour classes of var. Castillo; the differences were not significant. Data: Osorio Pérez et al.4

Mixed harvests. A 2017 study cupped defined mixes of ripe, overripe, semi-ripe and unripe fruit against the farm’s own harvest.6

Mix Caturra Colombia
M1: 100% ripe 82.98 82.07
M2: 80% ripe, 13% overripe, 7% semi-ripe 82.95 80.35
M3: 60% ripe, 26% overripe, 12% semi-ripe, 2% unripe 83.65 82.02
M4: farm harvest (control) 82.48 81.05

Source: Martínez et al.6 For Caturra the differences were not significant; for var. Colombia, M1 and M3 scored significantly higher than M2 and M4.6 The authors read the similarity of M1 and M3 as confirmation that less than 2.5% unripe fruit does not affect quality.6 They also cite earlier work in which, from 2.5% immature fruit upwards, 30% of cups were rejected.6

7880828486pointsAxis starts at 78Caturra · M1 100% ripe82.986 Caturra · M2 13% overripe,7% semi-ripe82.956 Caturra · M3 26% overripe,12% semi-ripe, 2% unripe83.656 Caturra · M4 farm harvest82.486 Colombia · M1 100% ripe82.076 Colombia · M2 13%overripe, 7% semi-ripe80.356 Colombia · M3 26%overripe, 12% semi-ripe,2% unripe82.026 Colombia · M4 farm harvest81.056
Data table
pointsSource
Caturra · M1 100% ripe82.986
Caturra · M2 13% overripe, 7% semi-ripe82.956
Caturra · M3 26% overripe, 12% semi-ripe, 2% unripe83.656
Caturra · M4 farm harvest82.486
Colombia · M1 100% ripe82.076
Colombia · M2 13% overripe, 7% semi-ripe80.356
Colombia · M3 26% overripe, 12% semi-ripe, 2% unripe82.026
Colombia · M4 farm harvest81.056
Cup scores of four harvest mixes in two varieties. All scores fall within about 3.3 points. Data: Martínez et al.6

Risks and defects arising at harvest#

Cenicafé links each class of faulty fruit to cup defects:7

  • Green (unripe) fruit – fermented, dirty, acrid and green cup defects.7
  • Overripe fruit – vinegary and fermented defects.7
  • Black fruit – acrid and carbonous defects.7

Delays and fermentation in sacks. Cherries held in bags or heaps heat up and ferment quickly.1 Leaving cherry more than 12 hours or overnight before pulping is one of the practices Cenicafé lists behind stained parchment, vinegary beans, moulds and ochratoxin A in wet parchment coffee.8

Mycotoxins. In a 2026 Colombian study (var. Castillo), ochratoxin A in the green coffee averaged 2.89, 2.98 and 3.03 µg/kg for red-orange, crimson and red-wine fruit processed without delay (the paper’s Table 2 gives 3.17 µg/kg for red-wine).9 Holding cherry for 24 or 48 hours at 15 or 20 °C gave a highest treatment mean of 3.35 µg/kg (3.31 µg/kg in the abstract), and neither holding time nor temperature had a significant effect.9 The authors conclude that the initial quality of the fruit matters most for bean safety.9 Fruit collected from the ground is a separate ochratoxin A risk.1

Reference protocol#

The table brings together the parameters reported in the sources. It describes practice and research findings, not a single correct recipe.

Parameter Value reported Source
Ripe colour Bright red to deep red, or yellow Haile & Kang2
Best harvest window Semi-ripe to overripe, 210–224 days after flowering (var. Colombia) Marín-López et al.3
Mature colour classes Red-orange, crimson and red-wine fruit cup the same Osorio Pérez et al.4
Green fruit in harvest Quality falls from 2.5% Cenicafé7, Martínez et al.6
Fruit on the ground Do not collect; same-day only in wet climates Codex1
Time to processing As soon as possible; same day; ≤ 10 h (Kenya); ≤ 6 h (Colombia) Codex1, FAO Lao PDR12, FAO Kenya13, Cenicafé7
Holding No bags or heaps; do not mix days’ harvests Codex1, FAO Lao PDR12
Hand sorting Remove immature, diseased, insect-damaged and dry fruit and debris FAO Kenya11
Flotation Skim off floaters: overripe and undeveloped fruit, sticks, leaves Haile & Kang2, Codex1

Measurement#

  • Colour. Research classifies ripeness by colour, for example by the red coordinate (a*) of a colorimeter.5
  • °Brix. A refractometer on fresh pulp. Values rise through ripening,3 but pulp sugar does not track bean sugar late in ripening.5
  • Firmness and detachment force. Measured with a penetrometer or dynamometer in research.3,6

Open questions#

  • A °Brix target. The studies opened here measured °Brix but published no picking threshold that ties it to cup quality. Numbers circulating in the trade could not be traced to a primary source and are left out.
  • Share of each harvest method by country. No primary source for these shares was found, so they are left out.
  • Optical sorting of cherry. No study measuring the effect of colour sorting of cherry on cup quality was opened for this entry.
  • Unripe fruit in flotation. Floaters include undeveloped fruit,2 but how well water separates sound unripe cherries from ripe ones was not measured in the sources here.

Key studies#

Marín-López et al., 2003 – Physical and chemical changes during ripening. Sampled var. Colombia weekly from 182 to 231 days after flowering and followed weight, firmness, soluble solids, pigments and cup quality. The best cups came from semi-ripe to overripe fruit; green and dry fruit gave poor ones.

Cenicafé 54(3), 208–225 · Cenicafé repository

Osorio Pérez et al., 2023 – Chemistry and cup of three mature stages. Compared red-orange, crimson and red-wine fruit of var. Castillo. Cup scores and most compounds did not differ; only fructose and glucose rose. The authors argue that dark fruit that is still turgid should not be called overripe.

Agronomy 13(2), 341 · doi:10.3390/agronomy13020341

Baptistella et al., 2024 – Sugars in fruit and bean. Split cherries into six red-colour classes and measured sugars in the skin-and-pulp, juice and bean separately. Sucrose nearly doubled in the outer fruit, but not in the bean, which suggests the two are regulated separately.

Plants 13(13), 1853 · doi:10.3390/plants13131853

Martínez et al., 2017 – Harvest composition and the cup. Cupped defined mixes of ripe, overripe, semi-ripe and unripe fruit in Caturra and Colombia. Scores stayed within a narrow band, and a mix with 2% unripe fruit matched pure ripe fruit for var. Colombia.

Vitae 24(1), 47–58 · doi:10.17533/udea.vitae.v24n1a06

Osorio et al., 2026 – Ripeness, delays and mycotoxins. Held cherries of three mature stages for 24 or 48 hours at 15 or 20 °C, and extended fermentation. Holding time and temperature had no significant effect on ochratoxin A, and the authors conclude that fruit quality at the start matters most.

Frontiers in Plant Science 17, 1734522 · doi:10.3389/fpls.2026.1734522

Contested
  • Are overripe cherries a defect? The Codex code of practice says overripe fruit should be removed before processing1, and a Cenicafé guide links overripe fruit to vinegary and fermented defects.7 But in a Cenicafé ripening study the best cups came from semi-ripe, ripe and overripe fruit (210–224 days after flowering)3, and a 2023 study found no difference in cup score between red-orange, crimson and red-wine fruit, arguing that dark fruit that has not lost its turgor should not be called overripe.4 The disagreement is partly one of definitions.
  • How much green fruit is tolerable: Cenicafé guidance puts the threshold at 2.5% green fruit7, and a 2017 study found no loss below it.6 In the same study, a mix with 13% overripe and 7% semi-ripe fruit lowered the score of one variety (Colombia) but not of the other (Caturra).6
  • How soon after picking: Cenicafé guidance sets a limit of 6 hours before pulping7, a Kenyan field report 10 hours13, and the FAO manual for Lao PDR the day of harvest.12 Codex asks only for processing as soon as possible.1
Share or cite this entry

Sources

  1. ACodex Alimentarius Commission (FAO/WHO) (2009). Code of Practice for the Prevention and Reduction of Ochratoxin A Contamination in Coffee (CAC/RCP 69-2009). Codex Alimentarius.
  2. BHaile, M.; Kang, W. H. (2019). The Harvest and Post-Harvest Management Practices' Impact on Coffee Quality. Coffee – Production and Research, IntechOpen. doi:10.5772/intechopen.89224 · other copies
  3. AMarín-López, S. M.; Arcila-Pulgarín, J.; Montoya-Restrepo, E. C.; Oliveros-Tascón, C. E. (2003). Cambios físicos y químicos durante la maduración del fruto de café (Coffea arabica L. var. Colombia). Cenicafé 54(3), 208–225.
  4. AOsorio Pérez, V.; Matallana Pérez, L. G.; Fernandez-Alduenda, M. R.; Alvarez Barreto, C. I.; Gallego Agudelo, C. P.; Montoya Restrepo, E. C. (2023). Chemical Composition and Sensory Quality of Coffee Fruits at Different Stages of Maturity. Agronomy 13(2), 341. doi:10.3390/agronomy13020341 · other copies
  5. ABaptistella, J. L. C.; Assoni, G.; da Silva, M. S.; Mazzafera, P. (2024). Variation in Soluble Sugars in Arabica Coffee Cherry Fruits. Plants 13(13), 1853. doi:10.3390/plants13131853 · other copies
  6. AMartínez C., V. M.; Aristizábal T., I. D.; Moreno C., E. L. (2017). Evaluation of the composition effect of harvested coffee in the organoleptic properties of coffee drink. Vitae 24(1), 47–58. doi:10.17533/udea.vitae.v24n1a06 · other copies
  7. APuerta-Quintero, G. I. (2001). Cómo garantizar la buena calidad de la bebida del café y evitar los defectos (Avances Técnicos 284). Cenicafé.
  8. APuerta-Quintero, G. I. (2008). Riesgos para la calidad por la comercialización de café pergamino húmedo (Avances Técnicos 373). Cenicafé.
  9. AOsorio, V.; Montoya, E. C.; Rayo-Mendez, L. M.; Harris, G. K. (2026). Impact of maturity stage and prolongation of post-harvest processing and mucilage fermentation time on mycotoxin levels in coffee. Frontiers in Plant Science 17, 1734522. doi:10.3389/fpls.2026.1734522 · other copies
  10. ATavares, T. O.; de Oliveira, B. R.; Silva, V. A.; da Silva, R. P.; dos Santos, A. F.; Okida, E. S. (2019). The times, movements and operational efficiency of mechanized coffee harvesting in sloped areas. PLOS ONE 14(5), e0217286. doi:10.1371/journal.pone.0217286 · other copies
  11. AMutua, J. (2000). Post Harvest Handling and Processing of Coffee in African Countries, Annex 5: Coffee Post Harvest Handling and processing in Kenya. FAO.
  12. AWinston, E.; Op de Laak, J.; Marsh, T.; Lempke, H.; Chapman, K. (2005). Arabica coffee manual for Lao-PDR. FAO Regional Office for Asia and the Pacific, Bangkok.
  13. AMutua, J. (2000). Post Harvest Handling and Processing of Coffee in African Countries, 3.0 Post Harvest Handling and Processing of Green Coffee in African Countries. FAO.

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”.

Suggest a change

On Harvesting and cherry sorting. Tell us what you would change and why. A source that supports it helps most.