Directive (EU) 2024/825 applies 27 September 2026.

See every fixed date
Eco Worldwide Sourced reference Claim checker

Does hemp use less water than cotton?

Short answer

Yes — but how much less depends entirely on which measure you use. A 2023 review of 28 published studies found hemp's crop water requirement 38% lower than cotton's, and its irrigated water footprint 91% lower. Hemp's advantage is mostly an irrigation advantage rather than a total-water one, and the figures repeated most often online come from a 2005 report whose own authors cautioned against generalising them.

The size of the gap depends on what you measure

Four water metrics, same comparison, same study. The spread between them is the whole story — and it is why two people can both cite real research and arrive at very different claims.

Hemp's water use versus cotton, by metric

Percentage lower than cotton. Wise et al. (2023), review of 28 published sources.

Crop water requirement38% lower
Water footprint60% lower
Crop irrigation requirement84% lower
Irrigated water footprint91% lower
0%100% lower than cotton

The two metrics at the bottom are irrigation measures. The two at the top include rainfall. That single distinction accounts for almost the entire spread, and it is the thing most coverage leaves out.

Green, blue and gray water are not the same thing

Green water
Rainfall stored in soil and used by the crop. It fell on that field whether or not anything was planted. High green water use is not, by itself, an environmental cost.
Blue water
Surface and groundwater abstracted for irrigation. This is the figure that bears on water stress, aquifer depletion and competition with drinking supply. It is the number that matters.
Gray or dilution water
A modelled volume of freshwater notionally required to dilute pollution from production to acceptable concentrations. A pollution metric expressed in water units — not water anyone actually withdraws.

Chapagain and colleagues put the global average water footprint of cotton at 3,644 m³ per tonne of seed cotton — split almost evenly between green water at 1,827 and blue at 1,818. For a tonne of finished cotton textile the figure is 9,359 m³, of which 4,443 is green and 4,917 blue. One cubic meter per tonne is one liter per kilogram, so that is roughly 9,400 litres per kilogram of finished textile, a little over half of it irrigation.

Why "cotton uses 10,000 litres per kilo" is misleading The figure is real, but it describes a global average for finished textile, roughly half of which is rainfall that fell on the field regardless. Quoted without that split, it implies ten thousand litres drawn from a river or aquifer for every kilogram. The actual irrigation component is closer to half that — still substantial, and a different claim.

Where the popular hemp figure comes from

Cherrett and colleagues at the Stockholm Environment Institute compared cotton, hemp and polyester in 2005. They put cotton at 9,788–9,958 litres per kilogram of lint, covering growth and processing. For hemp they reported 2,123 litres per kilogram of useful fiber using an experimental green decortication process, and 343 litres per kilogram using traditional processing.

Those two hemp figures differ by a factor of six, for the same crop, in the same report. The difference is entirely processing method.

The authors flagged their own hemp data The report states that the experimental results "might not be representative," because the green decortication method tested proved too costly. The authors also noted that hemp's seasonal precipitation requirement varies between 500 and 700 mm, and that their water calculations rest on assumptions about fiber losses in processing that may not reflect commercial operations. A caveat the original authors placed on their own number should not vanish when the number is repeated.

This does not make the finding wrong. The 2023 review, working across 28 sources rather than one experiment, points the same direction. It does mean that a specific figure carried forward for twenty years from a single trial of a process that was never commercially adopted deserves the qualification its authors gave it.

The three sources side by side

Note the functional units. They are not the same, and comparing across them directly is the most common error in this subject.

StudyYearFunctional unitCottonHemp
Chapagain, Hoekstra,
Savenije & Gautam
2005 per tonne seed cotton 3,644 m³
green 1,827 · blue 1,818
not studied
Chapagain, Hoekstra,
Savenije & Gautam
2005 per tonne finished textile 9,359 m³
green 4,443 · blue 4,917
not studied
Cherrett et al.
Stockholm Environment Institute
2005 per kg fiber / lint 9,788–9,958 L 2,123 L green decort.
343 L traditional
Wise, Baziotopoulos, Zhang,
Leaming, Shen & Selby-Pham
2023 relative, across 28 sources baseline −38% CWR · −60% WF
−84% CIR · −91% IRF
Two of these are not comparable Chapagain's 9,359 m³ per tonne is finished textile. Cherrett's 9,788–9,958 litres per kilogram is lint, before it becomes fabric. The numbers look almost identical and describe different stages of production. Setting them beside each other as agreement, or as a range, would be wrong.

Common questions

Q

Does hemp use less water than cotton?

Yes, but the size of the gap depends on the measure. A 2023 review of 28 published studies found hemp's crop water requirement 38% lower than cotton's, its water footprint 60% lower, its crop irrigation requirement 84% lower and its irrigated water footprint 91% lower. The advantage is mainly an irrigation advantage rather than a total-water one.

Q

How much water does cotton actually use?

Chapagain and colleagues put the global average at 3,644 cubic metres per tonne of seed cotton — 1,827 green and 1,818 blue — and 9,359 cubic metres per tonne of finished cotton textile, of which 4,443 is green and 4,917 blue. One cubic meter per tonne equals one liter per kilogram.

Q

What is the difference between green, blue and gray water?

Green water is rainfall stored in soil and used by the crop. Blue water is surface or groundwater abstracted for irrigation, and it is the figure that bears on water stress. Gray or dilution water is a modelled volume notionally needed to dilute pollution to acceptable concentrations — a pollution metric expressed in water units, not water anyone withdraws.

Q

Is it true cotton uses 10,000 litres per kilogram?

The figure is real but usually quoted misleadingly. It describes a global average for finished textile, roughly half of which is rainfall that fell on the field anyway. Quoted without the green and blue split it implies ten thousand litres drawn from a river or aquifer per kilogram; the irrigation component is closer to half that.

Q

Where does the claim that hemp uses half the water of cotton come from?

Largely from a 2005 Stockholm Environment Institute report by Cherrett and colleagues, which gave cotton at 9,788 to 9,958 litres per kilogram of lint and hemp at 2,123 litres using an experimental green decortication process, or 343 litres using traditional processing. The authors stated their experimental results might not be representative, because the method tested proved too costly.

Q

Does a lower water footprint make hemp the better fiber?

Not on its own. Land use, pesticide and fertiliser demand, processing chemistry, yield per hectare and fiber quality all bear on which crop suits a given product. Regional variation can also swamp crop choice — cotton's blue water requirement was 2,150 cubic metres per tonne in India against 3,860 in Pakistan, in similar climates.

What this does not settle

  • Where it is grown swamps what is grown. Chapagain's team found cotton's blue water requirement in India at 2,150 m³ per tonne against 3,860 in Pakistan — similar climates, an 80% gap, driven by irrigation infrastructure.
  • Hemp is far less studied than cotton. Cotton is one of the most heavily analyzed crops in agriculture. Hemp fiber has a fraction of the literature.
  • Water is one impact among several. Land use, pesticide load, fertiliser demand, processing chemistry, yield per hectare and fiber quality all bear on which crop is preferable.
  • Processing can dominate cultivation. The six-fold spread between the two hemp figures came entirely from how the fiber was separated, not how it was grown.
  • None of this describes a specific garment. These are crop and fiber averages.

Sources and method

  • Chapagain, A.K., Hoekstra, A.Y., Savenije, H.H.G. & Gautam, R. (2005). The water footprint of cotton consumption. Value of Water Research Report Series No. 18. waterfootprint.org
  • Cherrett, N. et al. (2005). Ecological footprint and water analysis of cotton, hemp and polyester. Stockholm Environment Institute. sei.org
  • Wise, K., Baziotopoulos, T., Zhang, Y., Leaming, S., Shen, D. & Selby-Pham, S. (2023). Comparative study of water requirements and water footprints of fiber crops hemp and cotton. Journal of Agrometeorology. agrimetassociation.org

Figures are taken from the primary studies, not from secondary coverage. Where a study's own authors placed a caveat on a number, the caveat is reproduced with it. Functional units are stated for every figure because the most common error in this subject is comparing across incompatible ones.

Components were checked against stated totals. In the Chapagain report the green and blue figures sum to one cubic meter above the stated total in both tables, which is rounding, and is noted here rather than silently reconciled.

Sources verified 13 September 2026.