Lab file · From the literature
How long should cold brew steep?
Short answer
Caffeine levels off around 6–7 hours at room temperature. Everything else keeps dissolving, and in one tasting, 14 hours beat 22.
- Caffeine and 3-CGA level off
- 6–7hSource [1]
- Extraction yield at 14 h (coarse, Huila)
- 15.21%Source [2]
- Extraction yield at 22 h (same coffee)
- 20.39%Source [2]
- Steep the tasters scored best
- 14hSource [2]
The data
Most cold brew recipes say 12 to 24 hours. A 2017 study timed it. Megan Fuller and Niny Rao steeped four versions of the same Kona coffee (two roasts, two grinds) at room temperature and sampled them all day. Caffeine and 3-CGA, the chlorogenic acid they tracked, rose fast for about three hours, slowed, and reached equilibrium between 6 and 7 hours in every sample [1].
The 24-hour readings came out a little higher than the 6.7-hour ones in every case, but the authors judged those gains not significant. Their wording: brewing near 400 minutes is “adequate to extract the majority” of both compounds. For the medium roast, medium grind, 3-CGA went from 480 to 510 mg/L across those extra 17 hours.
- 6.7 h (400 min)
- 24 h (1440 min)
Data table
Values in mg/L
| 6.7 h (400 min) | 24 h (1440 min) | |
|---|---|---|
| Medium roast, medium grind | 1,060 | 1,180 |
| Medium roast, coarse grind | 1,130 | 1,230 |
| Dark roast, medium grind | 970 | 1,080 |
| Dark roast, coarse grind | 930 | 990 |
Re-plotted from values reported in Table 2 of source [1]. Means of 6 measurements; the paper gives ±30 to ±90 mg/L 95% confidence intervals and calls the gain after 400 minutes not significant.
That’s caffeine. It isn’t the whole cup. A 2019 study from Colombia measured total dissolved solids (strength) and extraction yield (how much of the coffee ended up in the water) at 14 and 22 hours. Both kept rising. The coarse-ground Huila went from 15.21 % to 20.39 % extraction yield in those eight hours [2].
- 14 h
- 22 h
Data table
Values in %
| 14 h | 22 h | |
|---|---|---|
| Coarse grind, Huila | 1.54 | 2.04 |
| Coarse grind, Nariño | 1.61 | 1.96 |
| Medium grind, Huila | 0.71 | 0.94 |
| Medium grind, Nariño | 0.94 | 1.09 |
Re-plotted from values reported in Table 1 of source [2]. Brewed at 60 g coffee to 700 g water in a filter bag.
The grind moved those numbers even more than the clock did. That’s its own file: does grind size matter?
So the extra hours do something. Whether it’s something you want is a separate question. The same study had three trained tasters score every brew, and the best overall scores for both coffees went to coarse grind at 14 hours, not 22 [2]. The authors suggest that long steeps may pull out more of the compounds behind off notes. That’s their hypothesis. They didn’t measure it.
How they measured it
Fuller and Rao, 2017. 35 g of coffee in a tea filter bag, in 350 mL of filtered tap water, in a lidded Mason jar at 21–25 °C. They sampled every 15 minutes for the first hour, every 30 minutes to hour 7, hourly to hour 12, and once more at 24 hours. Caffeine and 3-CGA were measured by HPLC. Each coffee was brewed three times and each sample run twice. The full time course appears only as a figure in the paper. The values printed in a table are at 400 minutes (6.7 hours) and 1,440 minutes (24 hours), and those are the ones re-plotted above.
Cordoba and colleagues, 2019. 60 g of coffee to 700 g of water at 20 °C, grounds in a filter bag, steeped 14 or 22 hours. Two Colombian Caturra coffees (Huila and Nariño), medium roast, at a medium and a coarse grind. Dissolved solids were calculated from a °Brix reading, and extraction yield from that. Each brew was made twice and measured in triplicate. Three trained tasters scored the cold brews at 19 °C.
Limits
- Room temperature only. Both studies brewed at 20–25 °C. Fridge brewing is slower, and neither measured how much slower, so we don't know the fridge equivalent of 7 hours.
- The 6–7 hour plateau covers two molecules: caffeine and one chlorogenic acid. Flavor is hundreds of compounds, and their timing wasn't measured.
- The full extraction curve exists only as a figure in the 2017 paper. The printed values are 6.7 and 24 hours, so that's all we plot. The shape in between is theirs to show, not ours to trace.
- Nobody here measured total dissolved solids between 7 and 14 hours. The 2019 study starts at 14. There's a gap right where most home recipes live.
- The timing study used one coffee (a Kona Typica from one roaster) at one ratio. The tasting study used two Colombian coffees. Your beans may run faster or slower.
- 14 beating 22 hours comes from three tasters. Trained ones, but three. Treat it as a lead, not a law.
- Both studies steeped grounds in a filter bag without stirring. Loose grounds or a stir halfway may change the pace. Nobody tested it.
Takeaway
At room temperature, steep at least 7 hours so the caffeine is out. Around 14 hours is a sensible stop: it's what the one tasting panel preferred, and the extra hours mostly add dissolved solids. In the fridge, go longer. How much longer hasn't been measured.
Related gear and coffee
Illustration Toddy · Maker
Home Cold Brew System
Immersion · Felt · Mixed
Illustration OXO · Maker
Good Grips Cold Brew Coffee Maker
1183 mL · Immersion · Stainless mesh
Sources
- Fuller, M. & Rao, N. Z. (2017). The Effect of Time, Roasting Temperature, and Grind Size on Caffeine and Chlorogenic Acid Concentrations in Cold Brew Coffee. Scientific Reports 7, 17979. doi:10.1038/s41598-017-18247-4Peer-reviewed
- Cordoba, N., Pataquiva, L., Osorio, C., Moreno Moreno, F. L. & Ruiz, R. Y. (2019). Effect of grinding, extraction time and type of coffee on the physicochemical and flavour characteristics of cold brew coffee. Scientific Reports 9, 8440. doi:10.1038/s41598-019-44886-wPeer-reviewed
Charts are re-plotted from numbers printed in the sources' text and tables, never traced from their figures. How we pick and handle sources: about the Lab.