How Roasters Run a Cupping Session: SCA Protocol Step by Step

Every roastery that takes quality seriously ends up at the same table. Bowls in a row, a kettle on, spoons in a cup of hot water. Cupping is how you check a roast and judge a green coffee before you buy it. It also catches problems before they reach a customer's bag.

This guide walks you through how to run a cupping session the way a production roaster does. You'll get the coffee cupping protocol step by step and the numbers from the current SCA standard. You'll also get a simple routine for weekly QC cupping. The focus is the method, not the gear. If you're still putting your table together, the cupping lab equipment collection covers that side.

Key Takeaways

• Cupping is a fixed brewing method that strips out barista variables, so the coffee itself is the only thing you're judging.
• The SCA replaced its 2004 cupping protocol with the Coffee Value Assessment (CVA) in late 2024. SCA-102 now sets the sample prep and tasting mechanics.
• The dose is 8.25 g of coffee per 150 ml of cup capacity, so a 200 ml cup takes 11 g.
• Roast samples to a medium cupping level (about Agtron Gourmet 63) and rest them 8 to 24 hours.
• Pour at 93 ± 3°C, break the crust after 3 to 5 minutes, skim, and start tasting near 70°C.
• Cup every coffee on the table in parallel: all fragrance, then all pours, then all tasting rounds.
• A QC routine works best when it's boring: same day, same cups, same reference coffee every week.

Quick Answer: How Do You Run a Cupping Session?

Roast each sample to a medium cupping level and rest it 8 to 24 hours. Weigh 8.25 g of whole beans per 150 ml of cup volume, then grind each cup separately just before the session. Smell the dry grounds. Pour water at about 93°C to the rim and wait 3 to 5 minutes. Break the crust with three stirs while you smell the aroma, then skim the foam. Once the coffee cools to about 70°C, slurp each cup and note flavor, aftertaste, acidity, sweetness, and mouthfeel. Taste at least three times as it cools, then record your scores.

Why Roasters Cup

A café cups to choose a supplier. A roaster cups to run the business. Three jobs sit behind almost every session on a roastery table.

Quality control on production roasts

You pull a sample from each production batch and cup it against your reference. Did today's Colombia taste like last week's? A batch that drifts shows up here, sometimes even when the color reading looks normal.

Green coffee buying

Importers send pre-shipment and arrival samples. You sample roast them, cup them side by side, and decide which lots earn a contract. Arrival samples matter as much as offer samples. You want to know the coffee that landed is the coffee you approved.

Roast profile development

Change one variable on the roaster, then cup the result next to the old profile. Cupping is the only honest way to know whether a longer development time made the coffee sweeter or just flatter.

What Changed: The SCA Cupping Protocol in 2026

For about 20 years, most of the industry cupped to the 2004 SCAA protocol and its 100-point form. In late 2024 the Specialty Coffee Association adopted the Coffee Value Assessment as its official cupping standard. Three documents now govern the work:

• SCA-102 covers sample preparation and tasting mechanics. This is the part this guide follows.
• SCA-103 is the descriptive assessment. You rate how intense each attribute is and tick the flavor categories you find.
• SCA-104 is the affective assessment. You rate your impression of quality, which still produces a final cupping score.

The mechanics barely changed. A roaster trained on the old protocol will recognize every step. What changed is the paperwork, and a few numbers got a tolerance band instead of a single fixed value. Plenty of roasteries still run their own version of the 2004 form for internal QC, and that's fine. Just keep the mechanics consistent so your scores stay comparable week to week.

Preparing Samples: Roast Level and Resting Time

Bad sample roasts ruin good cuppings. If the roast is too dark, it hides defects. Too light, and the coffee tastes grassy and sour no matter how good the green is.

SCA-102 calls for a medium roast, defined as a ground reading of about 63 on the Agtron Gourmet scale. Most roasters check this on a color reader. Keep every sample on the same profile so you're comparing coffees, not roasts. Air cool the beans right away, with no water quench.

Rest matters too. Cup between 8 and 24 hours after roasting. Cup sooner and the coffee is still loaded with CO2, which muddies aroma and flavor. Wait much longer and the brightest notes start to fade. Store samples in sealed bags in a cool, dark spot while they rest.

Most roasteries roast these samples on a dedicated machine so the production roaster stays free. A 100 g to 300 g sample roaster lets you roast four or five lots in the time one production batch takes.

The SCA Ratio, Grind and Water

These are the numbers people get wrong most often, so here they are in one place.

One detail trips up a lot of new cuppers. The ratio is coffee to cup volume, not 8.25 g in every cup. Put 8.25 g in a 240 ml cup and you'll cup a weak brew all morning.

Use clean, odor-free, filtered water with no chlorine. Distilled water and heavily softened water both flatten the cup. Before each new sample, run a few grams of it through the grinder to purge the last coffee.

Enter your cup size and how many coffees you're cupping below. The planner works out the dose, the water and a run sheet for the table.

The Session Step by Step

Here's the full coffee cupping protocol in order. The one rule that holds it all together: work the whole table in parallel. Do step one for every coffee, then step two for every coffee, and so on. That keeps all the samples at the same temperature and stage.

1. Code the cups. Label the bottoms or use a random code so tasters can't see which coffee is which. Bias is real, even for experienced cuppers.
2. Grind and smell the fragrance. Grind each cup straight into its bowl. Lean in and smell the dry grounds. Note what you get, but don't score yet.
3. Pour. Start a timer and pour 93°C water to the rim of every cup, in the same order you'll taste them. A crust of grounds forms on top.
4. Smell the crust. While it steeps, smell the undisturbed crust on each cup. This is the first half of your aroma notes.
5. Break the crust. At 3 to 5 minutes, push through the crust with a spoon and stir about three times, nose close to the cup. The burst of aroma at the break is often the most useful moment of the whole session. Rinse your spoon in hot water before touching the next cup.
6. Skim. Once every crust is broken, use two spoons to lift the foam and floating grounds off each cup. Leave the settled grounds alone.
7. Taste in rounds. When the coffee reaches about 70°C, slurp a spoonful hard enough to spray it across your palate. Note flavor, aftertaste, acidity, sweetness, and mouthfeel. Spit into a spittoon cup so the caffeine doesn't blur your judgment by coffee four.
8. Go around again. Taste each coffee at least three times as it cools below body temperature. Defects and sweetness both show up more clearly as the cup cools.
9. Score, then talk. Finish your form before anyone shares notes. Discuss afterwards, and return to a cup if a colleague's comment changes how you read it.

A table of four to six coffees takes about an hour from grinding to final scores. Plan for the kettle and the grinding too. Those eat more time than people expect.

Scoring With a Cupping Form

Your form is a record, not a test. Its job is to let you compare today's coffee with last month's.

On the CVA forms, the descriptive side asks how strong each attribute is and which flavor groups you picked up. The SCA flavor wheel helps here when you're stuck between "fruity" and something more specific. The affective side asks how much you liked each attribute, and it rolls up into a final score.

The old 2004 form still shapes how the trade talks. A coffee at 80 points or above is called specialty. Under 80 is commercial grade. Q Graders calibrate to those lines, and importers price lots around them.

For internal QC, you don't need every box filled. Many roasters run a short sheet: pass or fail against the reference, plus a few flavor notes and one line on any defect. The important part is that every cupper on your team uses the same sheet the same way.

Running a QC Cupping Routine

A single great cupping tells you very little. A weekly one, run the same way for months, tells you everything. Here's a routine that holds up in a busy roastery:

• Fix a day and time. Morning, before anyone has had lunch or a second espresso.
• Keep a reference coffee. Freeze sealed portions of a batch you're happy with. Cup it alongside production samples so you're comparing against something real, not a memory.
• Cup every new production profile and every new green lot. Arrival samples go on the table before the lot goes into the roaster.
• Pair the cup with a number. Log the roast color reading next to each cup score. When a coffee tastes off, you can see whether the roast moved or the green did.
• Cap the table. SCA guidance puts a formal session at no more than six coffees. Past that, palates get tired and scores drift.
• Write it down where the roaster can see it. Notes in a drawer don't change the next roast.

The prep sheet below puts the routine in order, from the day before to the moment you pour. Print it and keep it by the kettle.

What We See Roasteries Add First

Most new roasteries start cupping with what they have: a scale, a burr grinder, a kettle, a few bowls and spoons. Our roastery equipment list treats that kit as day-one gear for good reason.

The first serious upgrade is usually not fancier bowls. Look at CoffeeTec's online orders for new lab equipment, January to October 2026. The QC instrument roasters buy most often is the Lighttells CM-100 Plus color analyzer, ahead of moisture meters. That tracks with the routine above. Once you cup every week, you want a number beside each score. A roast color reader gives you one in about a minute.

A sample roaster tends to come next, once green buying picks up. CoffeeTec's online orders for used sample roasters, January to October 2026, include Aillio, Ikawa, Mill City, Probat and Roest units. It isn't a new pattern either. Our client projects page lists a small Probat BRZ2 sample roaster we sold for Linea Caffe in San Francisco.

Complete roastery packages carry the same bench. The turnkey roastery business currently listed on CoffeeTec includes a San Franciscan SF-1 sample roaster and a lab grinder alongside its production roasters.

Common Cupping Mistakes

• Dosing by cup, not by volume. Weigh to the ratio for your cup size, and keep every cup identical.
• Cupping production roasts as if they were sample roasts. A dark espresso roast will taste roasty on any table. Judge it against its own reference, not against a green sample.
• Cupping the same day you roast. Give samples at least 8 hours to rest.
• Grinding everything an hour early. Fragrance fades fast. Grind as close to the pour as you can.
• Skipping the spoon rinse. One dirty spoon carries a defect from cup to cup.
• Talking during tasting. One "this one's fermented" and the whole table finds fermentation.
• Wearing scent to the table. Cologne, scented soap or a spicy lunch changes what everyone smells. Keep the room and the cuppers fragrance-free.
• Tasting once and scoring. Hot coffee hides a lot. The cool rounds are where defects and sweetness show up.

Selling Lab Equipment You've Outgrown?

Labs change as roasteries grow. A two-barrel sample roaster replaces a single, a benchtop analyzer replaces a handheld meter, and the old unit ends up on a shelf. Another roastery is often looking for exactly that piece. You can list your used QC and lab equipment with CoffeeTec, with no commission taken from your asking price. We write the listing, handle buyers, and arrange the logistics.