Skip to content
Tasting Cup

Try a city, a roaster, a process like “washed”, or a tasting note like “jasmine”.

Tasting Cup
Sign in

Yours

Brewing guide

Pour-Over Coffee Guide

A short and practical guide to pour-over coffee.

Although simple at first glance, the world of pour-over coffee is complex. Navigating the various options for brewers, filters, grinds, and brewing techniques can be overwhelming. This guide aims to demystify pour-overs by providing practical advice based on scientific research and real-world testing.

1Choose a brewer

To choose a brewer, you should first understand how its shape and brewing method change the cup, and second, consider your personal preference: the flavors you like, how hands-on you want to be, and the brands you trust.

Percolation or immersion

Pour-overs brew in one of two ways. Most let water run through the coffee as you pour. A few can also hold the water in, so the coffee steeps.

How percolation and immersion work and typically taste
TraitPercolationV60, Kalita, ChemexImmersionClever, Switch, French press
How it worksFresh water passes through the coffee and drains as you pour.Coffee steeps in the water until released, growing more concentrated.
ClaritySeparate, distinct notesBlended, integrated notes
AcidityOften brighterOften softer, rounder
SweetnessVivid, fruit-likeMellow, rounded
BodyLighter, cleanerFuller, most of all when using metal filters
Floral & fruit notesDefined, crisp, juicyLess distinct, sometimes jammy
Chocolate & nutty notesEasier to pick apartRich and cohesive
FinishClean, crispLingers longer

The filter matters as much as the method. In one lab study of a Kenyan coffee, paper filters (V60, AeroPress) gave more caramel and floral aromas, and metal filters (French press) more fruity and roasted ones. The French press scored lowest on every trait except acidity, where it matched the V60. The Clever and Switch use paper, so expect a smaller gap in body.

Cone or flat bottom

A cone narrows toward one opening, making a deeper coffee bed. A flat bottom spreads the same coffee across a wider, shallower base, which helps with even water distribution. Shape also decides how much water slips around the bed instead of through it, called bypass, which dilutes the cup.

Cone vs flat, in a lab

UC Davis researchers brewed the same coffees in drip machines fitted with either a cone or a flat-bottom basket, keeping the water, temperature and flow the same. Tasters could tell the two shapes apart, and the cone brewed stronger.

With a light roast, tasters picked out more

Flat
  • Dried fruit
  • Sweet
  • Floral
Cone
  • Citrus
  • Berry
  • Sour

Shape mostly mattered in combination with the roast: on its own it changed only a few traits (smoke aroma, sweetness and tobacco flavor). Consumers split into four taste groups, and no shape won with all of them.

Two cones, V60 vs Origami

0.02points of TDS apart, one pour

≈0.1points apart, two or three pours

TDS (total dissolved solids) is strength: the percentage of the cup that is dissolved coffee. Filter coffee usually lands between about 1.15% and 1.45%. On a 1.35% cup, 0.02 points is about 1.5% stronger, and 0.1 is about 7%, close to brewing at 1:15 instead of 1:16. In the UC Davis cone vs flat study, tasters needed a gap of about 0.24 points to notice a difference in strength, so both gaps here are likely too small to taste. The V60 read stronger with more pours. One person’s test. More on strength.

Popular brewers

60° cone, spiral ribs

Hario V60

One large hole. Grind and pour set the pace.

Filter
V60 cone
Largest brew
SizeCoffeeWater
01≈19–21 g320 ml
02≈35–40 g600 ml
03≈41–47 g700 ml

Our pick if You want the most control and the most recipes to copy.

Flat bottom

Kalita Wave

Three small holes limit drainage.

Largest brew
SizeCoffeeWater
155≈28–31 g470 ml
185≈42–47 g710 ml

Our pick if You want the brewer, not your pour, to limit flow.

Cone, 20 ribs

Origami Dripper

One large hole. Your paper changes the pace.

Filter
V60 or Kalita Wave
Largest brew
SizeCoffeeWater
S1–2 cups, no volume listed
M30 g500 ml

Our pick if You want cone and flat papers in one brewer.

Cone built into a carafe

Chemex Classic

Bonded paper against the glass wall.

Largest brew
SizeCoffeeWater
3 cup≈29–34 g500 ml
6 cup≈59–68 g1020 ml
8 cup≈79–91 g1350 ml
10 cup≈99–114 g1690 ml

Our pick if You brew for several people in one vessel.

V60-style cone with a valve

Hario Switch

Valve holds water, then releases it.

Filter
V60 cone
Largest brew
SizeCoffeeWater
02≈13–15 g230 ml
03≈21–24 g360 ml

Our pick if You want to move between steeping and percolation.

Immersion cone with a valve

Clever Dripper

Valve holds water until released.

Filter
#4 cone
Largest brew
SizeCoffeeWater
One size≈29–33 g500 ml

Our pick if You want a low-effort brew with little pouring technique.

≈ means estimate at 1:15 to 1:17 ratios.
In percolation, you pour in stages so the coffee bed is the real limit on amount. In an immersion brewer, all the water has to fit at once.

Other brewers

2Choose a filter

Every pour-over needs a filter, usually paper, sometimes cloth or metal, to keep the grounds out of your cup. Its material and details change the result, possibly more than the brewer’s shape does.

Flow rate

Flow rate is how fast water moves through the coffee, and it shapes the whole brew. Slower flow keeps water in contact with the grounds longer, so it extracts more. Grind size is the usual way to control it, but the filter, the brewer’s shape and your pour change it too.

A gooseneck kettle helps by letting you pour a little at a time and move the stream slowly and evenly over the bed. The exact pattern matters less than keeping the flow steady: an uneven pour mixes under- and over-extracted coffee in one cup. The last stage, as the remaining water drains, is the drawdown. An unusually long one often means the grind is too fine or the filter is clogged.

Water-only flow, six runs per filter. Faster means less resistance with no coffee present.

Water-only flow rate and thickness by filter
FilterFlow (ml/s)Thickness (µm)
V60 cloth
18.1
690
Hario unbleached
15.3
203
Hario tabbed, bleached
11.03
206
Chemex unbleached
9.82
167
Chemex bleached
7.23
210
Hario tabless, bleached
6.89
242
Cafec V60 (paper not stated)
5.79
207

“Thick” Chemex paper

Chemex paper is known for being thick and slow, and it’s easy to assume one causes the other. Measured, it’s barely thicker than Hario’s V60 paper, yet water runs through it a third slower.

210 µmChemex bleached sheet (Hario tabbed: 206 µm)

34%slower flow than Hario tabbed

So thickness alone doesn’t predict flow. The V60 cloth in the table is three times thicker than either paper and flows fastest of all. How open the material is likely matters more.

Coffee changes the order

The table times water alone. Once coffee is in the filter, fine particles can clog some papers more than others, so a ranking can change when you actually brew.

  • Sibarist far the fastest
  • Hario tabbed among the slowest

One roaster’s ranking from brewing with coffee; it published no times. With water alone, papers can look identical: a barista timed Cafec Abaca at 23 s and Hario at 24 s. Test a new paper with your own coffee.

Same brewer, new paper

The Origami takes both cone (V60) and flat (Wave) papers, so swapping only the paper shows its effect. Kurasu brewed the same recipe ten times with each.

+9 saverage brew time with Wave paper

The V60 paper brewed faster and read slightly stronger. Staff tasting the cups asked whether they were different beans; the V60 cups were fruitier and more defined. Individual brews varied widely, so treat the 9s as rough average.

Rinse the paper (optional)

Pouring hot water through the empty paper before brewing washes out some of its papery taste and preheats the brewer. It also presses the paper against the walls. Skip it if your paper tastes clean to you.

Unbleached paper rinsed with hot water

Paper sheds a little of its own material into hot water, which is where a papery taste comes from. Gagné poured hot distilled water through an empty unbleached Hario filter and measured how much paper washed out with each pour.

0.958 mgPour 1100 ml
0.293 mgPour 250 ml
None foundPour 350 ml

Milligrams of paper material in each pour of rinse water. After the first 150 ml no more was detectable. Bleached papers shed less.

Oils and filter material

Coffee oils carry flavor and, with the finest particles, much of a cup’s body. That is one reason espresso, pushed through a metal screen, feels heavier than pour-over. The filter decides how much oil reaches the cup: paper holds back most of it, cloth such as a nel drip filter less, and metal the least. Porous ceramic filters such as cerapotta are another option, but we found no measurements of how much oil they hold back.
The oils also carry cafestol, a compound that raises LDL cholesterol. This is well studied health effect, rather than just a question of flavor.

Cafestol in the cup

  • Paper filter, home12
  • Boiled, then cloth28
  • French press, percolator≈90
  • Boiled, unfiltered939

Median milligrams per liter, Swedish samples. Espresso samples reached 2,447 mg/L, but a shot is small: an older analysis found about 1 mg per espresso against 7 mg per cup of boiled coffee.

What it does

Cafestol raises cholesterol in step with how much you take in, and the rise fades after you stop.

+0.57 mmol/LLDL cholesterol (+22 mg/dL) after four weeks of 60 mg cafestol a day, in 10 men

+0.23–0.30 mmol/Ltotal cholesterol at six or more cups a day of boiled or plunger coffee, in a survey of 21,083 Norwegians

At the Swedish medians, 60 mg is about 0.7 L of French press, or 5 L of paper-filtered coffee. The trial used pure cafestol, and the survey shows a link, not a cause. Across 14 trials, filtered coffee barely moved cholesterol.

3Set your grind

Key terms
Grind size
How finely the beans are broken up. Finer particles extract faster and slow the water down, so grind changes both flavor and brew time. It is measured in microns (µm).
Fines
The smallest particles a grinder makes. They can clog the filter and slow the brew, which is why grinders with similar settings can behave differently.

Suggested coarseness ranges for different brewers.

Brewer settings

Comandante C40 MK4
V60
15–25
Pour-over
16–34

clicks

Timemore C2
V60
13–22
Pour-over
13–29

clicks

1Zpresso J-Max
V60
1.6.2–2.8.4
Pour-over
1.6.6–3.8.1

rotations

Baratza Encore
V60
7–18
Pour-over
7–28

dial

Fellow Ode Gen 2
V60
2.2–5.2
Pour-over
2.2–8.1

dial

Finer isn’t automatically better. In a blind V60 series on a Weber EG-1, setting 8.0 tasted best and settings below 6 fell off.

4Pick a pouring strategy

Key terms
Ratio
Water per gram of coffee, such as 1:16. A lower second number makes a stronger cup. Recipes here run from 1:15 to 1:18.
Agitation
Anything that stirs the coffee, such as swirling, stirring or a forceful pour. More usually raises extraction, but it can push fines toward the filter.
Bloom
A small first pour, usually two to three times the coffee’s weight, that wets all the grounds. It rests briefly before the main pours. This lets fresh grounds release the carbon dioxide trapped during roasting, so bubbles don’t push water away during the main pours and the bed wets more evenly. Fresher coffee and darker roasts bloom the most.

Water is shown by volume throughout, treating 1 g of water as 1 ml.

James Hoffmann, Ultimate V60

Hario V60
Coffee
30 g
Water
500 ml
Ratio
1:16.7
Temp
As hot as practical
Grind
Medium-fine
Agitation
Swirl, stir, swirl
Total
3:30
  1. At 0:00, pour 60 ml.
  2. At 0:45, pour 240 ml.
  3. At 1:15, pour 200 ml.

Tetsu Kasuya, 4:6

Hario V60
Coffee
20 g
Water
300 ml
Ratio
1:15
Temp
92 °C
Grind
Coarse
Agitation
None specified
Total
3:30
  1. At 0:00, pour 50 ml.
  2. At 0:45, pour 70 ml.
  3. At 1:30, pour 60 ml.
  4. At 2:10, pour 60 ml.
  5. At 2:40, pour 60 ml.

Clever, house recipe

Clever Dripper
Coffee
18.5 g
Water
310 ml
Ratio
1:16.75
Total
2:30
  1. At 0:00, pour 310 ml.

Tolo, flat-bottom recipe

Orea V4
Coffee
20 g
Water
360 ml
Ratio
1:18
Temp
93 °C
Grind
600–800 µm
Bloom
60 ml
Kasuya’s split is easy to experiment with. More water in the first 40% adds acidity, less adds sweetness. The last 60% sets strength.

5Water and temperature

Key terms
Temperature
Coffee can be brewed across a wide range of temperatures. Hotter water generally extracts more under otherwise identical conditions, and it flows slightly faster.

Serving: 58 °C to 66 °C. Maximized consumer acceptance.

Brewing: 87 °C to 93 °C. No appreciable taste impact at fixed strength and extraction.

Preheat ceramic. Preheating cut heat loss in a porcelain cone, and double-walled glass held heat best.

Water

Water contains dissolved minerals that can affect the taste of the coffee. Some people prefer making their own water by adding minerals to distilled water to control the flavor profile more precisely.

150mg/L total dissolved solids (75–250 acceptable)

≈68mg/L calcium hardness

40mg/L alkalinity

SCAA Water Standard, 2009.

6Balance grind, time and heat

These settings pull against each other. In the same brewer, grinding finer normally slows the brew, so pairing fine with fast takes different gear.

Finer, faster

Needs a high-flow paper. Sibarist sells its filters on faster drawdowns that open up new grind sizes.

Maker claim

Coarser, faster

Less extraction at the same contact time, so cups tend to run thin or sour.

Rule of thumb

Finer, slower

Yield climbs, but harshness can follow. Gagné suspects the pairing itself is the problem.

One brewer

Coarser, slower

Needs a way to hold water back. With a valved Pulsar, Gagné uses about 800 µm over 3:30–4:30.

One brewer

Immersion vs percolation

Fresh water keeps passing through
Water grows more concentrated

Gagné guesses extraction explains under half the taste gap, with fines and colloids doing the rest. No controlled taste test found. Try it on a Hario Switch.

Is cooler a fair trade?

91 °C

Gagné warns that falling below this tends toward sour, and brews as hot as his kettle allows. Coarser and cooler both reduce extraction, so pairing them works against you. At matched extraction, 87 °C and 93 °C tasted alike.

Does a lighter roast need longer?

Unproven

We found no test of it. Brewers more often raise the heat: Hoffmann favors hotter for light roasts, and Gagné suggests slightly cooler for dark. In 10-minute brews of one coffee, yield was highest at the lightest roast, so “harder to extract” may be too simple. Rao puts the peak at light-medium.

7Know what “good” measures

Key terms
Extraction
The share of the dry coffee that dissolves into your cup. Too little tends to taste sour, too much bitter or astringent. Many people land around 18–22%.
Strength (TDS)
How concentrated the cup is, measured as total dissolved solids. It is separate from extraction: a cup can be strong yet under-extracted.
1.01.21.41.61618202224Extraction (%)Strength (% TDS)Golden CupFive V60s, one brewer

Preferences vary

118consumers, 27 coffees, 2 preference groups

Tastes varied most with strength. The authors say a single “ideal” should be reconsidered.

Repeatability

3:04–3:10drawdown across five back-to-back brews

Tracking dose, water and time gets most people far without a refractometer.

8Kettle and scale

Gooseneck kettles

Scales

Sources and limits

All 42 sources and their limits
  1. Gagné, An In-Depth Analysis of Coffee Filters (2019). Water-only flow, thickness and taint. Filters may have changed since.
  2. Gagné, The Repeatability of Manual V60 Pour Overs (2019). Five brews by one experienced brewer on one coffee.
  3. Gagné, Why Can’t We Grind Coffee Finer for Pour Over? (2020). One brewer, one blind tasting.
  4. Gagné, Why Do Percolation and Immersion Coffee Taste So Different? (2019). One analyst’s reasoning. His estimate of the taste split is a guess.
  5. Gagné, Extraction Uniformity and Channeling (2019). One analyst’s modeling of flow through a coffee bed.
  6. Gagné, How to Brew Better Coffee with a V60 (2018). Personal practice and reasoning.
  7. Gagné, The Pulsar Dripper (2023). His own recipe. He receives royalties for his design work on this dripper.
  8. Santanatoglia et al., Discrimination of Filter Coffee Extraction Methods, Foods (2023). One Kenyan medium roast, four brewers, an expert panel. Paper versus metal filter separated the results more than the brew method did.
  9. Orrje, Fristedt and Rosqvist, Nutrition, Metabolism and Cardiovascular Diseases (2025). Cafestol measured in Swedish workplace and home brews. Few samples per method, so medians are rough. We read the abstract.
  10. Jee et al., American Journal of Epidemiology (2001). Meta-analysis of 14 randomized trials published before December 1998. We read the abstract.
  11. Urgert et al., American Journal of Clinical Nutrition (1997). Randomized crossover trial, 10 healthy men, about 60 mg of pure cafestol a day for 28 days. We read the abstract.
  12. Urgert and Katan, Journal of the Royal Society of Medicine (1996). Review. Reports the effects are transient after people stop. We read the abstract.
  13. Svatun et al., Open Heart (2022). Tromsø Study, 21,083 adults aged 40 and over. Cross-sectional, so it shows association, not cause. We read the abstract.
  14. Gross, Jaccaud and Huggett, Food and Chemical Toxicology (1997). Diterpenes per cup by brew method. We read the abstract.
  15. cerapotta ceramic coffee filter. Maker’s site. It publishes no measurements of what the ceramic holds back.
  16. Batali, Ristenpart and Guinard, Scientific Reports (2020). Drip brewed. We read the abstract only.
  17. Cotter, Batali, Ristenpart and Guinard, Journal of Food Science (2021). Mostly college-age consumers. We read the abstract and a conference poster.
  18. Ristenpart, Cotter and Guinard, Scientific Reports (2022). Reanalysis of the Cotter dataset, read through a summary.
  19. Frost, Ristenpart and Guinard, Journal of Food Science (2019). Automatic drip brewers, 45 tasters for the difference test, 85 consumers. We read the abstract and Tolo’s summary; the light-roast flavor split comes from the summary.
  20. Scientific Reports, Caffeine content in filter coffee brews (2024). Yield by roast from 10-minute brews. Read through a search excerpt.
  21. Rao, The twin myths of “easier to extract” and “overextraction”. Practitioner review. Read through a search excerpt.
  22. Tolo, Flat-bottom vs cone dripper. Retailer article.
  23. Kurasu, Origami with V60 vs Wave filters. Retailer test: Origami S, one recipe, ten brews per paper, staff tasting. Results ranged widely.
  24. Everyday Beans, V60 vs Origami. One person’s test.
  25. Home Ground Coffee Roasters, Qualities of Filter Papers. Ranks filters but publishes no times.
  26. Chowhound, Best Paper Coffee Filters. One competition barista’s water-only timing.
  27. Honest Coffee Guide, grind size chart. Published starting ranges, not measurements of your grinder.
  28. Barista Hustle, Temperature’s Hidden Effect. Paywalled. We read a search excerpt only.
  29. James Hoffmann’s Ultimate V60, on Hario USA. Published recipe.
  30. Tetsu Kasuya’s 4:6 method, on Hario Europe. Published recipe.
  31. Clever Dripper product page. Maker’s practical capacity (500 ml) and recipe (18.5 g to 310 g), read 9 November 2026.
  32. Chemex Classic Series support page. Maker’s specifications, read 9 November 2026. Each size makes “up to the ounce amount” of coffee, so we add back the water the grounds keep.
  33. Kalita Wave series page. Maker’s page, in Japanese. Read 9 November 2026.
  34. Hario V60-02 Switch, on Hario Canada. Distributor listing of the maker’s specifications.
  35. Hario USA, SWITCH Immersion Dripper. Hario’s listed capacities. We matched them to sizes using the product numbers. The 02 is listed as finished coffee, so we add back the water the grounds keep.
  36. Rogue Wave Coffee, Hario V60 listing. Retailer’s practical capacity for each size.
  37. Your Dream Coffee, Hario V60 sizes. Independent blog with affiliate links. It puts the 03 nearer 1,000 ml.
  38. Seattle Coffee Gear, Kalita Wave guide. Retailer guide. It gives both 24 and 26 oz for the 185; we use the lower figure. It calls these yields, but we treat them as water in, which matches published 185 recipes such as 42 g to 700 ml.
  39. Slow Pour Supply, Origami Dripper. Retailer listing: 1–2 cups (S), 1–4 cups and up to 30 g of coffee (M). The M’s volume is our estimate at a 1:16.7 ratio.
  40. Our dose estimate. Makers list volumes or cups, not doses. We divide each listed volume by 15 and 17, the usual filter ratio range. Where a volume is finished coffee, we first add back about 2 g of water per gram of grounds.
  41. Sibarist Cone product page. Maker’s own claims, not independently tested.
  42. SCA Golden Cup and SCAA Water Standard (2009). Values taken from published excerpts of the standards.

Log something

Sign in or create an account

Add private notes to beans and shops, and keep track of your bags and brews.

No password to remember. You will receive a login code via email.