What is a soil wetting agent, and does your garden need one?

Wetting agents · the complete guide

By Joe, Founder of Dr Forest · August 2026 · 20-minute read

The physics is straightforward. The trial evidence is messier than the packaging suggests, and rather more interesting for it.

A wetting agent is a surfactant that lowers the surface tension of water so it soaks into soil evenly instead of beading up and running off. That is the entire function. A bag of compost gone dusty in the shed, a hanging basket that empties out of its drainage holes in four seconds, a lawn shedding water on a dry August afternoon: in all three cases the soil or substrate has turned water-repellent, and a wetting agent is the thing that gets water back in contact with the surfaces it needs to wet. It feeds nothing and it holds nothing. It changes how water and a dry particle behave when they meet.

This page is the reference piece for the whole set, and the short answers to the common questions sit in the wetting agent FAQ. It covers the mechanism, the product types, what the peer-reviewed trials show, application rates, and where the evidence runs out. Eight shorter pages go deeper on individual jobs, and they are listed further down. If the ground you are treating is grass, start with the lawn wetting agent page instead.

Why soil stops taking water in

Soil that will not accept water is rarely short of pore space. The pores are still there. What has changed is the chemistry of the surfaces lining them.

Organic matter in soil contains amphiphilic molecules: waxes, fatty acids, fungal residues and the breakdown products of roots and leaves. Amphiphilic means one end of the molecule has an affinity for water and the other end does not. While the soil is moist, those molecules sit with their water-liking ends facing outwards. Dry the soil hard enough and they turn round, presenting the water-avoiding end to the pore space. The particle is now coated in something closer to a candle than a sponge.

The cleanest demonstration of this is old. Ma'shum and Farmer, working on a severely water-repellent sandy pasture soil in South Australia in 1985, found that freeze-drying converted it into a readily wettable soil, and that rewetting followed by oven-drying regenerated the repellency [1]. The coating was still present throughout. What changed was its orientation. Mao and colleagues reviewed the same phenomenon from nanoscale to ecosystem scale in 2019 and described repeated wetting and drying as the thing that shifts the water-particle relationship, leaving a soil hydrophilic when wet and hydrophobic when dry [2].

That reversibility is the practical point for a gardener. Repellency is a state your soil moves into, and it can move back out.

Where it happens in UK gardens

Retail copy tends to file this as a sandy-soil problem. The best British dataset says otherwise. Doerr and colleagues sampled 41 common soil and land-use types across humid temperate Britain at 0–5, 10–15 and 20–25 cm depth, using water drop penetration time for persistence and critical surface tension for degree [3]. They found repellency in surface samples of all major soil textural types at most permanently vegetated sites, with levels at the worst reaching those of the most severely affected areas anywhere in the world. Tilled sites were virtually unaffected. Above a water content of about 28% by volume, repellency was absent.

Two useful things follow. Cultivation is protective, which is part of why a dug vegetable bed behaves itself whilst an established lawn or an undisturbed border does not. It is also how a dry patch on a lawn gets started. And moisture is the switch. Repellency also decreased with depth and showed no consistent relationship with texture or organic matter, so you cannot predict it from a soil type alone.

Compost, peat-free and otherwise

Container substrate is where most UK gardeners meet this first. Durand and colleagues measured contact angles on white milled peat by capillary rise and recorded 90.0° at both 40% and 50% moisture content by weight, easing to 87.0° at 60%, and treat 40% by weight as the lower limit below which a peat has very little ability to rewet [4]. Peat-free mixes are not exempt, though wood fibre and coir both come out of the same tests rather better than peat does [5]. Dried-out compost is a job of its own, and it has pages of its own: how much to use in pots and containers, and why pots and baskets stop taking water.

What a surfactant actually does

Pure water at 20 °C has a surface tension of 72.74 millinewtons per metre, with a stated uncertainty of 0.36 [6]. That figure is why water beads. Molecules at the surface are pulled inwards by their neighbours and the surface behaves like a stretched skin.

A surfactant molecule is built the same way as the organic coatings causing the trouble: water-liking head, water-avoiding tail. Put enough of them into solution and they crowd the air-water interface with their tails outwards, which breaks that skin up. They also adsorb onto the particle surfaces themselves, which is the half of the story most explanations leave out.

Wetting agent, surfactant and adjuvant all describe that same molecule doing slightly different jobs, and the places where the three words part company are set out in wetting agent, surfactant or adjuvant.

Karagunduz, Pennell and Young measured both effects on Ottawa sand and an Appling soil using the laboratory surfactant Triton X-100 [7]. Surface tension fell from 72 to 32 mN/m. The solid-liquid contact angle fell from 40° to 10°. Both changes stopped at the critical micelle concentration, which for that surfactant is 0.15 g per litre. Above it the extra molecules form micelles in the bulk solution and no further change happens at the interface. Adding more past that point buys you nothing.

Contact angle is what decides whether water enters a pore under its own steam. Capillary rise scales with surface tension multiplied by the cosine of the contact angle. At 90° the cosine is zero and there is no spontaneous capillary rise at all, so water has to be pushed in. Durand and colleagues state exactly that for peat sitting at or above 90°. Bring the angle down to 40° and the cosine is 0.77. Bring it to 10° and it is 0.98.

Here is the complication that marketing tends to skip. Lowering surface tension improves the contact angle and weakens the capillary pull at the same time, because that pull scales with surface tension too. Karagunduz and colleagues measured the net effect and found soil water contents decreased incrementally as surfactant concentration rose towards the critical micelle concentration. A wetting agent improves the entry of water into soil. In a clean physical system it slightly worsens how much water that soil holds at a given suction. Any copy claiming a wetting agent helps soil hold more water is arguing with that measurement.

One further finding shapes how you apply the stuff. Ogunmokun and Wallach compared surfactant solutions poured onto hydrophobic columns against plain water poured onto columns already pre-treated with surfactant [8]. Pre-treatment won clearly. Their conclusion is that the limiting step is surfactant molecules diffusing to and adsorbing onto the hydrophobic surfaces, rather than the surface tension of whatever liquid you are pouring. That is the mechanistic reason behind the standard advice to apply, allow some dwell time, then water in.

Granular, liquid and natural

Three formats reach UK gardeners, and they differ more in how the dose arrives than in the underlying chemistry: granules that dissolve off a dry carrier over days and suit open ground, liquid concentrates that act on contact and go exactly where you put them, and plant-derived saponin from soap nut pericarp. Saponin is the weaker surfactant of the two chemistries, with surface-tension minima of about 52 and 47 mN/m against 32 to 37 mN/m for a good synthetic, and hard water raises the dose it needs by roughly eight to ten times [11] [12]. Household washing-up liquid is the usual substitute and a poor one; the case against washing-up liquid sits with the rest of the homemade wetting agent options. Which format suits which job, and how the natural or synthetic chemistries actually compare, is worked through on a page of its own.

One finding from that literature is worth carrying away, because it runs against my own commercial interest. O'Brien and colleagues compared the penetrant and retainer categories printed on the front of bottles across sand-based putting greens in Arkansas and Texas, and through 2018 and 2019 found no significant differences in volumetric water content between them on any date or at any depth; in 2021 the season-long difference averaged less than 2.5% volumetric water content, and one of those differences ran the wrong way [13]. Their recommendation is that products be classified on performance data rather than terminology, and they have described the two words as effectively marketing terms.

What the evidence actually shows

Read only the product pages and you would think the chain of cause is settled: wetting agent goes on, soil holds more water, plant does better. The field data does not line up that tidily, and the way it fails to line up is the most interesting thing in this article.

Dempsey, Fidanza and Kostka reported five site-years of a soil surfactant on amenity turf, one in Pennsylvania and four in Ireland, across native loam, USGA-specification sand and natural links sand, in randomised complete block designs with four replications [14]. Turfgrass quality was significantly better in the treated plots in every single site-year. Volumetric water content was not significantly different in any of them, and in two of the five it sat marginally lower in the treated plots. Their own conclusion says it plainly: no significant or consistent differences in water content, and significantly enhanced quality.

Chang and colleagues ran the nearest thing to a lawn study anyone has published: St Augustinegrass on a Boonville fine sandy loam at 3.7% slope, two 21-week seasons, four applications a year at 0.9 g per square metre of active ingredient, four replicates [15]. Turf quality came out at 7.1 against 6.8 untreated in the first year and 8.0 against 7.6 in the second, with green cover of 91% against 85% in 2016. Soil moisture was similar to untreated at every fertiliser level in both years. There was no significant effect on runoff volume, nitrate export, phosphorus loss or water drop penetration time. The abstract of that paper describes soil moisture as slightly improved whilst the results text reports it as similar to untreated, which is a discrepancy worth knowing about if you go and read it.

Aamlid and Pettersen pushed further in the other direction. Monthly surfactant on sand-based creeping bentgrass greens at Landvik in Norway decreased the average water content of the surface 7.5 cm from 0.193 to 0.168 m³ per m³ in 2014, with a similar result the following year [16]. Water penetration time fell. Quality improved only under deficit irrigation and only in 2015, one of two years and one of three irrigation regimes. The soil got drier and the surface accepted water faster, both at once.

So what is the mechanism connecting product to plant? The likeliest answer is in the title of Soldat, Lowery and Kussow: surfactants increase the uniformity of soil water content and reduce repellency on sand-based greens [17]. Their treated plots, irrigated at only 30% of potential evapotranspiration, held visual quality below a well-irrigated control and above a drought-stressed one. Water goes into more of the rootzone instead of running down a handful of preferential channels and leaving the soil between them dry. Average moisture can hold flat, or fall, whilst the plant does better, because roots do not grow in the average. The honest framing for all of this is "more even, not more".

Does your garden actually need one?

I would do this before buying anything, mine included. A wetting agent has no job to do on soil that already takes water, and the test for that costs nothing.

Go to the spot that is actually giving you trouble, on a dry day, and test the surface rather than a spadeful from underneath, because repellency lives in the top few centimetres and fades with depth. Put three or four separate drops of water on it, the size that comes off a fingertip, and watch them. Drops that vanish in a second or two mean the contact angle is already well below a right angle, the soil is wetting normally, and nothing in a bottle will improve on that. Drops that hold their shape as beads mean the surface has turned repellent, and the physics above is the reason: at 90° the cosine is zero, there is no spontaneous capillary rise at all, and water has to be pushed in rather than pulled. Drops still sitting there after a minute are the case a wetting agent is actually for.

Two honest limits on what that tells you. The test is blunt: Bachmann and colleagues found the water drop penetration method sensitive only between 85° and 115°, inside a range that runs from 0° to 142° [20]. It will tell you that a surface repels water; it will not tell you how badly. And repellency is a moisture state rather than a soil type. That British survey found it absent above about 28% water by volume [3], so a bed that beads in August may take water perfectly well in October, and testing the day after rain will tell you nothing at all. Test it dry, which is when it is a problem.

Rates

Dr Forest Natural Wetting Agent is diluted at 0.5 to 2 teaspoons per litre depending on the job, drenched or sprayed on, then watered in. A teaspoon is 5 ml, so 1 teaspoon per litre is 5 ml per litre, which is 0.5% by volume. Sizes are 500 ml at £14.00, 1 litre at £22.00 and 5 litre at £80.00. Other brands differ, sometimes by a lot, so read the label you have rather than this one. The rate for each individual job, and how much to use per square metre, have a page of their own, and the wetting agent calculator will do the coverage arithmetic for a given area.

One number I am not going to give you is the in-use concentration of saponin in milligrams per litre. The saponin content of the concentrate is not published, so that arithmetic cannot be done honestly. Apply at label rate and do not over-concentrate on the theory that more will work harder; the physics section above explains why it will not.

On watering in, there is a directly relevant trial. Powlen and Bigelow compared immediate post-application irrigation against delaying it by roughly 18 hours, over nine applications at 14-day intervals on a sand-based putting green [18]. Both surfactants improved seasonal quality either way, so the sky does not fall if you leave it, but watering in immediately gave better soil moisture and water penetration outcomes in late summer, which is when it matters most.

What a wetting agent will not do

This is the part I would want to read before buying, so here it is with the numbers attached.

The effect is shallow. Dekker and colleagues applied a commercial surfactant twelve times to a water-repellent grass-covered dune sand between 22 April and 23 November, and took 4,950 samples [19]. The soil was severely to extremely water repellent to more than 0.50 m depth during dry periods. Reduced repellency appeared in the surface 0.05 m. Below that, in the authors' words, no effects were observed, which they attribute to the surfactant adsorbing in the surface layer before it can travel further. They also state that the surfactant did not equalise the uneven moisture distribution below that surface layer. There was a real gain inside the top 5 cm: the critical soil water content was lowered distinctly, meaning the treated layer could dry further before turning repellent again, and the thatch layer at 0 to 0.025 m often held slightly higher water contents.

Repellency comes back. Song and colleagues dried and rewetted treated sand columns repeatedly at 55 °C. Performance degraded across three cycles, one product still ran at 34 mm per minute after the third, and another had stopped infiltrating entirely after the second [10]. Barton and Colmer recorded repellency ranging from 0.4 to 4.3 on the molarity-of-ethanol-droplet scale across a single irrigation season [9]. Every field programme in the evidence I have read reapplied every 14 to 30 days through the growing season. Nothing supports a one-application fix, and any product sold on that basis is overpromising.

The usual measurement is weak. Bachmann and colleagues compared methods across 24 soils and found that advancing contact angles from Wilhelmy plate and modified capillary rise agreed well from 0° to 142°, whilst the water drop penetration test was only sensitive between 85° and 115° [20]. Most before-and-after drop tests in marketing material are working inside a 30° window of a 142° range.

There is no UK domestic-lawn trial. I went looking and did not find one. The verified British material is an occurrence survey with no surfactant treatment in it [3], plus work on fungal contributions to repellency on golf greens. The nearest field trials to a British garden are Irish [14] and Dutch [19], both on managed turf with calibrated professional equipment. Nobody has published a trial of a consumer wetting agent on domestic borders, raised beds or ornamental containers.

And the one that lands closest to home. No peer-reviewed field or glasshouse trial of saponin as a soil wetting agent, for water-repellent soils or for container substrates, has been published. The turf and soil surfactant literature is about alkyl polyglucosides and ethylene-oxide/propylene-oxide block copolymers. The surface chemistry of soap nut saponins is well measured [11], and the mechanism is the same amphiphilic one that makes any surfactant work. The use of saponin specifically as a soil wetting agent rests on that mechanism and on vendor material, mine included, rather than on a published trial. I would rather write that down than let you assume otherwise.

One related result belongs here too. In laboratory agar bioassays, an isolated hederagenin saponin from soap nut pulp inhibited red clover seedling root growth with an IC50 of 16.64 mg per litre [21]. Those tests used germinating seeds in agar rather than established plants in soil, where sorption onto particles and rapid degradation cut the exposure considerably. It is still a good reason to work to the label rate rather than eyeballing a capful.

Final word

The threshold matters more than the bottle. That UK survey found repellency absent above about 28% water by volume, which is another way of saying the cheapest wetting agent is not letting the soil get that dry in the first place. When you have gone past it, on a lawn in August or a basket that has dried to dust, mix at label rate, wet the surface, give it half an hour, then water it in with about 5 mm. Judge whether it worked by watching the water rather than the colour: the first thing to check is whether water now sinks in where it previously beaded and ran off to the edges.

How often you repeat it depends on what you have treated. A pot or container that has dried right out usually takes two or three waterings with the solution before it wets evenly again, and then plain water until it next dries out hard; on borders, let the water decide, and treat again when it starts beading on the surface instead of soaking away. Turf runs on a monthly programme for three to four months where the dry patch is established, which is set out on the lawn wetting agent page. That monthly rhythm follows RHS guidance on dry patch, where the instruction to repeat once a month for three to four months belongs to the cultural programme of spiking and aeration that a wetting agent accompanies, rather than to its wetting agent line [22]. The monthly-versus-fortnightly question has been looked at directly by the University of Arkansas Agricultural Experiment Station, whose wetting agent studies compared application rates and timings; those are preliminary results on sand-based putting greens rather than domestic lawns, and the station's own conclusion is that “optimizing application rates and timings is product-specific, and there is not a one-size-fits-all strategy that works best across all wetting agents” [23].

Frequently asked questions