The Chemistry of Roasting and Caffeine Stability
Caffeine (1,3,7-trimethylxanthine) is a thermally stable alkaloid. It doesn't vaporize or meaningfully break down at the temperatures used in drum roasting — 180–230°C. Studies measuring caffeine concentration by dry bean weight across roast levels consistently find differences of roughly 1–3%, well within the noise of most lab analyses. A landmark study in the Journal of Medicinal Food (Sánchez-González et al.) confirmed what roasters already knew empirically: light-roasted Arabica measured fractionally higher in caffeine per gram than dark-roasted beans from the same lot, but the gap was negligible.
What does change dramatically is bean density. Green coffee contains roughly 10–12% moisture. During roasting, that moisture drives off and bean cell walls expand through pyrolysis — a thermal decomposition that destroys organic polymers and releases gases. A City roast (light-medium) loses about 14–16% of its original mass; a French roast loses 20–25%. The beans are physically larger and more porous after dark roasting, but each cubic centimeter of bean contains less actual material — and therefore less caffeine per unit of volume.
This is why the volume-versus-weight question dominates any honest comparison. Most home brewers measure with tablespoons or scoops. A scoop of dark roast beans, being less dense, holds fewer grams than the identical scoop of light roast beans. The result: brew the same "two scoops" across roast levels and you'll pour more caffeine from the lighter bag, simply because you scooped more mass.
First Crack, Second Crack, and the Roasting Stages
The roasting process progresses through identifiable thermal events that specialty roasters use as control points. Understanding these stages clarifies why caffeine stability is almost beside the point — the real flavor and density transformations happen before the bean ever reaches second crack.
Drying stage (100–150°C): Free moisture evaporates. Beans yellow and emit grassy aromas. Almost no flavor compounds are formed yet. Roasters target a controlled, even drying to avoid "baked" flavors caused by moisture trapped inside dense beans.
Maillard window (160–195°C): The pivotal phase. Reducing sugars and amino acids react across hundreds of simultaneous pathways, generating pyrazines, furans, and aldehydes responsible for the nutty, caramel, and sweet notes common to medium roasts. This is the developmental heart of coffee flavor — push through it too fast and the cup is flat; linger too long and it goes bitter before reaching first crack.
First crack (196–205°C) is the signature event of light roasting. Steam and CO₂ burst through cell walls with an audible snap. At this point, Maillard reactions have already produced hundreds of aroma compounds — the clean florals, fruit acids, and delicate tea-like notes characteristic of washed Ethiopians or Kenyan SL28. Specialty roasters targeting these flavor profiles pull the bean within 30–60 seconds of first crack.
Development time ratio (DTR) — the time between first crack and the drop — is one of the most important roaster parameters. A 20% DTR on a 10-minute total roast means 2 minutes of post-crack development. Short DTRs preserve acidity and brightness; longer DTRs build body and reduce perceived sharpness.
Second crack (225–230°C) marks dark territory. Cell walls fracture more deeply, oils migrate to the bean surface, and Strecker degradation replaces Maillard complexity with smoky, bitter, roasted-grain notes. The bean's original origin character — the terroir signal of altitude, variety, and processing — is largely overwritten at this stage. Beyond second crack into Italian or Spanish roast territory, carbonization begins. The bean is genuinely burned at the molecular level.
Roast Level × Caffeine: The Numbers
These figures represent approximate values based on published analytical chemistry of Arabica beans from single lots roasted to multiple targets. The Agtron score measures roast color via near-infrared reflectance — lower numbers indicate darker roasts.
| Roast Stage | Agtron Score | Mass Loss | Caffeine by Weight | Caffeine by Volume |
|---|---|---|---|---|
| Light (City) | 75–85 | ~14% | 100% (baseline) | ~120% |
| Medium (Full City) | 55–65 | ~17% | ~99% | ~110% |
| Medium-Dark (Vienna) | 45–55 | ~19% | ~98% | ~104% |
| Dark (French) | 35–45 | ~22% | ~97% | ~95% |
| Very Dark (Italian) | 25–35 | ~24% | ~96% | ~88% |
The caffeine-by-weight column confirms the chemistry: negligible differences, all within 4%. The caffeine-by-volume column tells the story that actually matters in most kitchens: the density drop between a City and an Italian roast creates a 25–30% caffeine gap when beans are measured by the scoop.
Species Matters More Than Roast Color
The Arabica / Robusta distinction produces caffeine swings that no roast adjustment can match. Arabica (Coffea arabica) typically carries 1.2–1.5% caffeine by dry weight; Robusta (Coffea canephora) ranges from 2.0–2.7%. A dark-roasted Robusta espresso will contain significantly more caffeine than a light-roasted single-origin Arabica from the same dose weight — not because of the roast, but because of genetics.
Canephora evolved at lower altitudes in central and western Africa where caffeine's insecticidal properties offered a meaningful competitive advantage. High caffeine is a defense mechanism, not a quality attribute — which partly explains why commodity Robusta blends have historically been used to cut espresso blends for cafes that prioritize stimulant delivery over cup clarity.
This becomes practically relevant when reading espresso blend labels. A "bold" blend listing Robusta near the top of its composition will deliver materially more caffeine than a 100% Arabica specialty single origin, even if the specialty bean is a light City roast and the blend is a dark Italian roast. If you want to manage your caffeine intake, species identification matters far more than roast color.
Brewing Method: The Largest Caffeine Variable
Even with dose weight and bean species controlled, brewing method introduces the widest real-world caffeine range. Contact time, temperature, and extraction efficiency vary dramatically across brewing formats.
| Brew Method | Typical Caffeine per Serving | Notes |
|---|---|---|
| Cold brew concentrate (diluted 1:1) | 200–300mg / 240ml | Extended steep maximizes extraction |
| Drip / pour-over | 90–150mg / 240ml | Standard household reference |
| French press | 100–140mg / 240ml | Full immersion, coarser grind |
| Moka pot (90ml) | 105–130mg | Near-espresso pressure |
| Double espresso (60ml) | 120–160mg | Typical specialty-café dose |
| AeroPress (standard recipe) | 90–130mg | Highly variable by recipe |
| Single espresso (30ml) | 60–80mg | Concentrated but small volume |
| Instant coffee (200ml) | 60–90mg | Lower extraction efficiency |
Espresso's reputation for "strongest coffee" comes from concentration per milliliter, not total caffeine per serving. A standard mug of drip coffee will typically deliver more total caffeine than a single espresso shot. Cold brew — prepared as an 18-hour steep at a 1:4 coffee-to-water ratio — flips the expectation entirely; by serving time, it may be the most caffeinated beverage in your kitchen.
Water temperature also modulates extraction. Water at 93–96°C (near boiling, not at it) optimizes caffeine solvation. Cold water extracts caffeine more slowly, but the extended brew time of cold brew more than compensates, achieving full extraction by hour 12.
Flavor Profiles at Each Roast Stage
Caffeine's near-invariance across roasts focuses attention back where it belongs: on flavor. The roasting arc from light to dark is fundamentally a journey from origin transparency to roast expression.
Light roast maximizes terroir. Washed processing and high-altitude growing produce fragile, volatile aromatics — bergamot, jasmine, citric acid, stone fruit — that survive only at lower drum temperatures. Ethiopian Yirgacheffe, Kenyan Nyeri, and Colombian Huila are typically best at City to City+ roasts, where the fruit acid profile remains intact and the sweetness is still green-apple-crisp rather than caramel-thick.
Medium roast finds the balance. Caramelization deepens body, adds hazelnut and toffee tones, and rounds out sharp acidity. This is the workhorse range for structured pour-overs, flat whites, and filter brews designed to appeal broadly without demanding single-origin literacy from the drinker.
Dark roast serves the espresso tradition, particularly the Italian and Portuguese models where robustness, low perceived acidity, and integration with milk are the design goals. Bitterness from caffeoylquinic acids and phenolic Strecker products becomes deliberate — assertive in a ristretto, smoothed by milk in a latte. The original bean variety is largely irrelevant at this roast level; the roast is the flavor.
The practical upshot: choose your roast for flavor, not for caffeine management. The caffeine difference between identical weight doses of light and dark roast is within the margin of measurement error in most home kitchens.
Frequently Asked Questions
Does dark roast coffee have more caffeine?
No. When measured by equal weight, dark roast coffee has marginally less caffeine than light roast — roasting degrades a small percentage of caffeine at higher temperatures, and the longer drum exposure of dark roasting removes slightly more. The widespread belief that dark equals more caffeine likely comes from measuring by volume (scoops): darker beans are less dense, so the same scoop holds fewer grams and less total caffeine.
Which type of coffee bean has the most caffeine?
Species is the dominant variable. Robusta (Coffea canephora) contains 2.0–2.7% caffeine by dry weight, roughly double the 1.2–1.5% of Arabica. No roast level adjustment to an Arabica bean can approach Robusta's baseline caffeine.
Does espresso have more caffeine than drip coffee?
Per ounce, yes — espresso is far more concentrated. Per serving, typically no. A single shot (30ml / ~70mg) contains less caffeine than a mug of drip coffee (240ml / ~120–150mg). A double shot is roughly comparable to one mug of drip.
What brewing method produces the most caffeine?
Cold brew prepared at a 1:4 coffee-to-water ratio and steeped 12–24 hours typically extracts the most caffeine per serving, due to the extended contact time compensating for cold-water's slower solvation. Serving concentration matters, however — a diluted cold brew is not automatically stronger than drip.
Should I measure coffee by weight or volume for consistent caffeine?
Always by weight. A gram of coffee at any roast level contains essentially the same caffeine, so weighing your dose eliminates the density variable that makes dark-roast scoops seem weaker. A digital kitchen scale is the single most effective tool for consistent, repeatable brewing across roast levels.
Conclusion
The roast-level-caffeine relationship is a case study in how a plausible intuition — darker means stronger — collapses under measurement. Bean density, not roast chemistry, is what makes a tablespoon of light roast deliver more caffeine than the same tablespoon of French roast. Species origin creates the largest real caffeine differences, brewing method creates the second largest, and roast level comes in a distant third with an effect too small to matter in practice. Choose your roast for flavor: light for origin transparency, medium for balance, dark for espresso tradition. Weigh your doses for consistency. Browse our roasted coffee selection — every bag notes the roast level and origin so you can match the flavor profile to your preferred brew method and dial in your ideal cup.