A cupping lab uses cups that hold 240 mL to the rim. Under SCA-102, what dose of coffee should go into each cup?

13.20 g ± 0.2 g. SCA-102 sets 8.25 g of coffee per 150 mL of vessel capacity, or 0.055 g per mL. For 240 mL, 240 × 0.055 = 13.20 g, weighed per cup with a ±0.2 g tolerance.

How does SCA-102 define cupping-level roast?

Ground L* 26–29 (about tile #65), or an Agtron Gourmet reading of 63. SCA-102 defines cupping-level roast as ground-coffee L* 26–29, roughly tile #65 on the Agtron/SCA Roast Color Classification System, or a target reading of 63 on the Agtron Gourmet scale.

What grind does SCA-102 specify for cupping?

70–75% passing a US No. 20 sieve, each cup ground separately. SCA-102 requires each cup to be ground separately so that 70–75% of the grounds pass a US No. 20 sieve with an 850 µm aperture, slightly coarser than typical paper-filter drip.

SCA-102 Sample Preparation: Roast Level, Dos | Free Guide 2026

The written exam covers cupping procedures "including setup" in detail, and three practicals depend on correctly prepared samples. The governing document is SCA Standard 102-2024, Coffee Value Assessment: Sample Preparation and Tasting Mechanics (SCA-102). Learn its numbers exactly.

What SCA-102 replaces

SCA-102 states that it supersedes these heritage SCA standards:

Coffee to Water Ratio for Cupping
Cupping Water Temperature
Cupping Water
Grind for Cupping
Roast for Cupping and Roast Level for Cupping
Cupping Room Size, Cupping Spoons, and Cupping Tables
SCA Cupping Protocol (2004 and any later editions)

It is designed for green Coffea arabica and may be adapted to other states of coffee and other Coffea species. It covers mechanics — preparing and tasting — while the rating rules sit in SCA-103 (descriptive) and SCA-104 (affective).

Roasting the sample

Roast level. The roast for specialty cupping should be what is usually described as medium. SCA-102 defines cupping-level roast as:

Ground-coffee lightness L* of 26–29 in CIELAB coordinates, corresponding approximately to Agtron/SCA Roast Color Classification System tile #65; or
A target reading of 63 on the Agtron "Gourmet" scale, or its equivalent on other roast meters.

Roast level should be controlled by colorimetry or infrared spectrometry. Dry-matter loss or bean volume increase may be used if they are correlated with color. When using roast meters, control grind level and coffee temperature; samples with a lot of chaff show larger error. If a coffee is cupped at another roast level, report the deviation to all parties.

The SCA's Q Grader preparation page expresses the same target as ground Agtron 63, about 58 whole bean.

Roasting problems. Roasts that miss the target range, or that roasted too fast (excessive heating rate) or too slow (insufficient heating rate, giving a "baked" bean), can show underdeveloped, burnt, or baked character. SCA-102 says such batches should not be used for cupping and should be repeated. Chapter 9 covers these problems in detail.

Consistency, cooling, and rest.

Keep all roasting parameters as constant as possible from sample to sample.
Air-cool immediately, without water quenching.
At about 20°C, store samples in airtight containers or non-permeable bags in a cool, dark place.
Rest 8 to 24 hours after roasting. If cupping cannot happen in that window, take steps to minimize flavor degradation.

Cups and dose

Vessels. Tempered glass or ceramic, 200–350 mL, top diameter 75–90 mm. All cups must be identical in volume, dimensions, and material.

Dose. Measure the cup's volume to the rim, for example by weighing the room-temperature water it holds and using 1 g ≈ 1 mL. Then dose 8.25 g of coffee per 150 mL of vessel capacity. Weigh the beans for each cup separately on a scale accurate to 0.1 g; because of bean size, a tolerance of ±0.2 g applies.

$\text{dose (g)} = \text{cup capacity (mL)} \times \frac{8.25}{150} = \text{cup capacity} \times 0.055$

Cup capacity to rim	Dose	Acceptable range (±0.2 g)
200 mL	11.00 g	10.80–11.20 g
240 mL	13.20 g	13.00–13.40 g
250 mL	13.75 g	13.55–13.95 g
300 mL	16.50 g	16.30–16.70 g
350 mL	19.25 g	19.05–19.45 g

Dosing 8.25 g into a 240 mL cup would make a much weaker brew than the standard intends, which is why the ratio, not the 8.25 g figure alone, is what matters.

Grinding

Grind as close to cupping time as practical.
Grind each cup separately.
Grind so that 70–75% of the grounds pass a US standard No. 20 sieve (850 µm). This is slightly coarser than typical paper-filter drip.
Set the grinder before the session. Before each new sample, grind a small amount of that sample to displace grounds left from the previous coffee.

How many cups per sample

Assessment	Method	Cups or brew
Affective (SCA-104)	Cupping required	5 cups per sample, so uniformity can be judged
Descriptive (SCA-103)	Cupping or batch brew	3–5 cups if cupping; if batch brewing, 55–60 g per liter (about 17:1 to 18:1 water to coffee)

When batch brewing for descriptive work, keep a small amount of dry grounds aside for fragrance assessment. Use a grind specific to the brewing method, for example a French press grind.

Setup checklist

Confirm roast color and rest time for every sample.
Measure cup capacity and calculate the dose.
Weigh whole beans for each cup separately (±0.2 g).
Purge the grinder with each new sample; grind each cup separately.
Arrange the coded cups so that all coffees can be brewed and assessed in parallel (Section 6.2).
Check that the water meets the SCA-102 water specification (Section 6.3).

SCA / CQI Q Grader Certification (Coffee)

6.1 SCA-102 Sample Preparation: Roast Level, Dose, Cups & Grind

Key Takeaways