SCA Coffee Sensory and Cupping Handbook
Sources: Coffee Sensory and Cupping Handbook by Fernández-Alduenda & Giuliano (SCA, 2021)
Authors: Mario Roberto Fernández-Alduenda and Peter Giuliano Publisher: Specialty Coffee Association, September 2021 Edition: No. 01 ISBN: 978-1-3999-0329-5
The handbook bridges sensory science and coffee industry cupping practice. It is based on empirical evidence and peer-reviewed research, cited by its authors as the most heavily referenced publication on coffee sensory science to date. It has four parts: Foundations of Sensory Science, The Coffee Sensory Experience, Applying Sensory Assessment to Coffee, and Coffee Cupping.
Part 01: Foundations of Sensory Science
History of cupping
Clarence E. Bickford (San Francisco, 1800s) invented systematic cup-testing, discovering that high-altitude Central American coffees were superior despite their small bean size — breaking the visual-grade-equals-quality paradigm. Ted Lingle’s 1984 Coffee Cuppers’ Handbook first systematized cupping science for the specialty trade and introduced a proto-lexicon of ~175 sensory terms. The SCA 100-point cupping protocol was developed in 1999, inspired by the wine industry. The Coffee Taster’s Flavor Wheel (2016) was updated using the WCR Sensory Lexicon — a scientific collaboration between SCA, WCR, and UC Davis.
Objectivity and subjectivity
A central distinction in this handbook: analytical sensory attributes (taste intensity, body level, acidity intensity) are considered objective even when individual tasters disagree. Value judgments (cupping score, grade, liking, preference) are subjective. These should never be mixed in a single test. This maps directly onto the distinction between descriptive (objective) and affective (subjective) sensory analyses. See Sensory Testing Methods.
Sources of bias and error
Bias and error in cupping are pervasive and significant. Key categories:
- Halo effect / response correlation: Studies show ~90% correlation among all cupping attribute scores — a coffee perceived as high quality gets scored high across the board regardless of actual attribute intensity
- Expectation bias: Knowing roast level, origin, or variety biases the score — double-blinding best practice
- Dumping effect: A conspicuous attribute causes the taster to record it under a wrong category (e.g., very astringent coffee causes tasters to mark astringency under “acidity”)
- Order / contrast effects: A defective sample makes the next sample score higher than warranted
- Social / authority bias: Cuppers influenced by others’ opinions — silence until all forms submitted is standard
- Sensory fatigue: Olfactory and taste receptor saturation; rest between samples; avoid “strong coffee” for at least one hour prior
Best practices to minimize bias: double-blind sample coding, randomized sample order, standardized cup type, controlled environment, cupper calibration sessions.
Part 02: The Coffee Sensory Experience
Five stages of coffee perception (Ch. 5)
- Fragrance: dry grounds — most volatile (lowest molecular weight) compounds; sweet aromatics most prominent
- Aroma: brewed coffee — heavier compounds volatilized by hot water; Maillard reaction products (caramel, nutty, cocoa) predominate; aroma character is distinct from fragrance
- Flavor: when sipping — retronasal olfaction combines with taste and touch; perceived as a unified “flavor image” in the brain
- Aftertaste: residual compounds coating tongue and pharynx, especially lipids; aromatic character persistent; length and character are independent quality attributes
- Evolution as coffee cools: flavors change dramatically, similar to wine “opening up”; due to changing volatility and temperature-dependent receptor sensitivity
Olfactory dimension (Ch. 6)
Two olfactory pathways — orthonasal (sniffing in, fragrance/aroma) and retronasal (breathing out from mouth during and after swallowing) — produce different perceptual qualities. Most of what we perceive as coffee “flavor” is retronasal olfaction. Blocking your nose eliminates nearly all flavor. Human beings excel at retronasal smell relative to other animals.
SCA-specific terminology (must be used precisely):
- “Fragrance” = smell of dry coffee grounds
- “Aroma” = smell of brewed coffee
~1,000 odor compounds in roasted coffee; only ~13–38 are “potent odorants” (active at trace concentrations). Sulfur-containing compounds are the most intense potent odorants — unpleasant in isolation, but central to the recognizable “coffee” character in combination. Maillard reaction produces dominant potent odorant families: furanones (caramel-like), pyrazines (roasty/earthy), phenolic compounds, damascenone, vanillin.
Smell memory is closely linked to emotion and craving (via the limbic system). A coffee’s aroma memory can trigger wanting/craving behavior — the neurological basis for café aromas driving consumer behavior. See Sensory Science.
Taste dimension (Ch. 7)
Five basic tastes active in coffee: sour (primary), bitter (primary), sweet (crossmodal — see below), salt (minor), umami (minor, under-researched).
Acidity: 10–12 major acids responsible; dominated by the chlorogenic acid family, carboxylic acids, phosphoric acid. Sourness predominates in light roasts; bitterness takes over in dark roasts. Acidity character depends on genetics, terroir, processing (washed > natural for brightness), roast, and brewing. “Juicy” acidity is a prized market attribute. See Sensory Attributes and Value.
Bitterness: commonly misattributed to caffeine. Caffeine accounts for only 10–20% of coffee bitterness. Primary contributors are:
- Chlorogenic acid lactones (up to 30% of bitterness): formed during roasting from chlorogenic acid breakdown
- Phenylindanes (up to 30% of bitterness): formed at higher roast temperatures from lactone breakdown; also studied for potential neuroprotective properties (inhibit amyloid-beta and tau aggregation — emerging research, not established)
- Quinic acid and caffeic acid (minor contributors, also from chlorogenic acid breakdown)
Bitterness increases with roast level. Light roasts have the least bitterness; very dark roasts are dominated by it. See Roasting.
Sweetness in coffee is a crossmodal phenomenon — not from dissolved sugars. No sweet compound in a normal brewed coffee reaches taste threshold concentration. Coffee’s perceived sweetness comes from sweet-smelling aromatics (vanillin, caramel compounds, fruit esters) perceived retronasally, creating a “crossmodal” sweet taste impression. This has direct implications for cupping: scoring “sweetness” is really scoring sweet aromatic presence, not dissolved sugar. See Cupping (Cata).
Saltiness: Light-roasted C. arabica is measurably saltier than light-roasted C. canephora. High-mineral water (high K⁺ and Na⁺) increases perceived saltiness.
Cup balance: The harmonious interaction of flavor, aftertaste, acidity, and body. No universal “balanced cup” — culturally and individually defined. Brew strength (TDS) and extraction % can be varied to shift balance. See Coffee Brewing Control Chart.
Tactile dimension — body and mouthfeel (Ch. 8)
Body = tactile expression in mouth: thickness (viscosity, from suspended polysaccharides) + texture (smoothness to roughness, from particle size). These are tactile, not taste, perceptions.
Astringency = mouth-drying sensation caused by chlorogenic acids and quinic acid binding to salivary proteins. Associated with unripe beans, over-roasting, and over-extraction. Undesirable in specialty coffee.
Body is affected by: species (C. canephora = lower lipids → thicker, rougher body), processing method (natural = more polysaccharides → thicker body than washed), roast level, brewing method (paper filter removes oils and particles → thinnest body; French Press → thicker; espresso → emulsified lipids, thickest).
Crossmodal effects (Ch. 9)
Inputs from non-taste senses measurably alter flavor perception — even in expert tasters. This is neuroanatomical, not just psychological, and cannot be trained away.
Key practical findings:
- Cup color: pink/red cups enhance sweetness perception; white cups enhance acidity perception
- Cup shape: angular shapes associated with bitterness and acidity; round shapes with sweetness
- “Bouba-Kiki” effect: applies to taste — round sounds/shapes → sweet; angular → bitter/sour
For cupping: standardize cup type, shape, and color across all sessions; never compare scores from different cup types or environments.
For product/service design: deliberately choose serving vessels to enhance desired sensory characteristics. Pink or red cups for naturally processed coffees → perceived sweeter. Green/yellow packaging for high-acidity coffees → expected sour/acidic experience.
Common language (Ch. 10)
WCR Sensory Lexicon: 110 scientifically validated attributes. Each attribute has: (1) a clear definition, (2) a physical sensory reference (food or chemical at a specific preparation), (3) an intensity value on a 15-point scale. Enables global tasters to share the same definition AND sensory experience.
Coffee Taster’s Flavor Wheel (2016): 9 primary → 16 secondary → ~80+ specific attributes, arranged by perceptual similarity (consensus of trained tasters). How to use:
- “Center-out”: identify primary category, then narrow to specifics
- Spatial proximity = perceptual similarity — use to resolve taster discrepancies (disagree on “blackberry” vs. “raspberry”? → both are “berry” → agree at that level)
Warning: using fruits to describe acidity without specifying “acidity” is poor communication practice. “Apple” or “lemon” as descriptors may be misunderstood as flavor notes rather than acidity character.
Part 03: Applying Sensory Assessment to Coffee
Three categories of sensory tests (Ch. 11–14)
| Type | Question | Example tools |
|---|---|---|
| Difference test | ”Are these two coffees detectably different?” | Triangulation, 3-AFC |
| Affective test | ”How much do consumers like this?“ | 9-point hedonic scale, JAR |
| Descriptive test | ”What is this coffee’s sensory character?” | CATA, descriptive panel |
See Sensory Testing Methods for full detail.
Affective testing (Ch. 13)
“Quality” in specialty coffee is culturally defined and subjective. Mediterranean cultures historically prize phenolic/bitter coffee. Third Wave/specialty markets prize light, fruity, clean cups. Both are valid preference systems. This means “quality” scores from different cultures are not directly comparable.
Consumer segmentation via preference mapping reveals that coffee consumers cluster into distinct groups with different preference drivers. Research on brew strength shows two consumer clusters: “strong coffee likers” (nutty, roasted, dark chocolate attributes drive liking) vs. “weak coffee likers” (tea/floral, sweet, cereal attributes drive liking). Kaiserblick’s specialty light roast targets the second cluster.
Descriptive analysis (Ch. 14)
Descriptive cupping (rapid method developed by Fernández-Alduenda): combines trained cuppers’ descriptive skills with the SCA cupping protocol. Dramatically reduces training cost vs. a full sensory panel while providing meaningful descriptive data. Use the nine innermost circle descriptors of the Flavor Wheel as a CATA ballot.
Part 04: Coffee Cupping
Sensory attributes and market value (Ch. 15)
Research from Cup of Excellence auction data (Traore, Wilson & Fields) links specific sensory attributes to price premiums:
| Attribute | Effect on value |
|---|---|
| Floral, sweet, fruity, spice aromas | Significant price premium |
| ”Off” aromas (rubber, overripe, musty, petroleum) | Price penalty |
| High acidity / “juicy” acidity | Higher price; correlated with high altitude |
| Flavor complexity (number of attributes detected) | Adds value |
| Creamy body, smooth mouthfeel, round body | Add price value |
| Balance | Adds value in single-origin context |
| Sweetness, uniformity, cleanness | Absence disqualifies specialty status |
| Sensory defects | Disqualify specialty grade |
See Sensory Attributes and Value for full analysis and Kaiserblick implications.
What cupping is and isn’t (Ch. 16)
Cupping is designed for green coffee evaluation in the trading context. It is not the optimal method for evaluating roasted coffee for a specific brewing method. A roaster evaluating an espresso blend should taste it as espresso, not as a cupping. Results may be misleading otherwise.
Uses of cupping along the value chain: discovery, quality screening at dry mill, lot creation (blending small lots into exportable mill marks), grading, pre-shipment sampling, arrival sampling, purchase sampling, quality feedback upstream.
Cupping panel size (Ch. 17)
| Panel size | Use case |
|---|---|
| 1 cupper | Exploratory only |
| 3 cuppers | Minimum for standard deviation calculation |
| 6 cuppers | Minimum for Q Grader-level reliable decisions (SCA research confirmed optimal) |
| 8–12 cuppers | Best for descriptive sensory research and competitions |
| >12 cuppers | Competitions and auctions; individual experience valued over accuracy |
SCA Cupping Protocol (Ch. 18)
Sample preparation:
- Roast: 8–12 min completion; Agtron Gourmet 63.0 or Commercial 48.0 (± 1.0); rest at least 8 hours, cup within 24 hours
- Ratio: 8.25g coffee per 150mL water (mid-point of the Gold Cup)
- Grind: slightly coarser than drip filter; 70–75% pass US #20 sieve; grind immediately before cupping, no more than 15 min prior
- Water: clean, odor-free; 125–175 ppm TDS (ideal); 90–96°C (195–205°F) at pour; pour to rim
Evaluation sequence:
- Fragrance (dry): within 15 min of grinding; lift lid, shake grounds gently, smell
- Infuse: pour to rim; steep 3–5 min undisturbed
- Aroma (wet crust): assess crust intact, then break — stir 3 times, let foam run down, sniff during break
- Flavor, Aftertaste, Acidity, Body, Balance: when cooled to ~71°C (160°F), ~8–10 min post-pour; slurp from spoon
- Sweetness, Uniformity, Clean Cup: as approaches room temperature ~38°C (100°F); 2 points per cup, max 10 per attribute (CATA-style: each of 5 cups independently assessed)
- Overall (Cupper’s Points): subjective quality impression
- Stop evaluating at 21°C (70°F)
Scoring:
- Scale: 6.00–10.00 in 0.25-point increments; 17-point scale
- Specialty coffee threshold: ≥ 80 points total after defect deductions
- Defects: taint = −2 pts, fault = −4 pts, multiplied by number of affected cups
- Uniformity, sweetness, clean cup: 2 pts/cup × 5 cups = max 10 pts each
Twelve evaluated categories: Fragrance/Aroma, Flavor, Aftertaste, Acidity, Body, Balance, Uniformity, Clean Cup, Sweetness, Overall, Defects (taint), Defects (fault).
Applications of cupping (Ch. 19)
- Single cupper margin of error: ±1–2 points — never rely on one cupper for contract-critical decisions
- Pre-shipment sample retained alongside delivery; arrival sample cupped against pre-ship sample to detect transit damage — standard practice for export contracts
- Trading context: cupping forms must be legible, clear, use common Flavor Wheel language, professionally written — may serve as contract documentation
Open Questions
- Nordic brewing standard: The SCA Brewing Handbook cited a 60–70 g/liter NCC standard (2011); current SCA standards may differ — verify before citing to customers
- Phenylindane neuroprotective properties: Emerging research (Mancini et al., 2018) suggests phenylindanes inhibit amyloid-beta and tau aggregation. Not yet established clinical fact
- CoE price premium data: Research (Traore et al., 2018) reflects a specific period of CoE auction data; may not fully generalize to current Salvadoran micro-lot market