How long should the sprinkler run? The question sounds like it needs a rule of thumb and it does not — it needs three numbers you can measure and one you can look up. Area and depth give litres directly, because one millimetre of water over one square metre is exactly one litre. Divide by the flow rate and by the share of that water that actually reaches the soil, and you have the minutes. The part most people miss is the second half: if those minutes exceed what the soil can absorb in one go, the honest answer is not one long run but several short ones with gaps, and this lays that out on a clock so you can see what the programme actually occupies.
Water, flow and clock
One millimetre over one square metre is exactly one litre.
The ground the sprinkler actually covers, not the whole plot.
12 to 15 mm for established lawn, 15 to 20 for a vegetable bed, 5 for new seeding.
Measure it: run into a bucket and time it. Guessing this is the usual source of error.
Early morning is best: low wind, low evaporation, and the foliage dries quickly.
15 minutes on heavy or compacted soil, 30 on open sand, 10 on a slope.
Result
Why depth is the natural unit for watering
Watering is discussed in millimetres for one very good reason: one millimetre of water over one square metre is exactly one litre. That makes the conversion between a depth — which is what plants experience and what soil holds — and a volume — which is what a meter bills and a tank holds — a matter of multiplying by the area and nothing else. No conversion factors, no density, no temperature correction.
It also makes the advice portable. 'Twelve millimetres twice a week' means the same thing to someone watering a two-square-metre border and someone watering a two-hundred-square-metre lawn, whereas 'forty litres' does not. Whenever you read a watering recommendation that is not in millimetres, it is worth converting before you act on it, because the millimetre figure carries the information about how deep the water is going, and the litre figure does not.
Efficiency is not a detail, it is a quarter of the answer
A sprinkler throws water into the air. Some of it evaporates before it lands, some is carried off by wind, some lands on the path, the wall or the foliage rather than the soil, and some lands so finely that it drifts. The share that ends up where you wanted it is the efficiency, and for a typical sprinkler it is somewhere between sixty and eighty percent. Drip and trickle systems, which put water directly onto the soil, run at around ninety.
This matters because it divides, and dividing by a number less than one is what turns a comfortable estimate into an under-watered garden. Nine hundred litres needed at seventy five percent efficiency is twelve hundred litres that has to come out of the tap — three hundred litres, or a third again, of water that has to be delivered for no benefit other than the wetting of a patio.
The soil has a speed limit, and it is slower than a sprinkler
Here is the constraint that makes the second half of this calculation necessary. Soil can only accept water so fast: a heavy clay might take eight to twelve millimetres an hour once the surface is wet, a compacted lawn less, an open sandy loam considerably more. A sprinkler delivers water much faster than that — twelve litres a minute over seventy-six square metres is about nine millimetres an hour, and over a smaller area it is far more.
When the application rate exceeds the infiltration rate, the surplus does not wait; it runs. It runs to the low corner, onto the path, down the drive, and it takes the top few millimetres of soil with it. You can recognise the result: a lawn that is wet at the edges and dry at the roots, and a water bill that does not match how the plants look.
Running the same total in two or three passes with twenty-minute gaps lets each application disappear into the soil before the next arrives. The total water is identical; the amount that reaches the roots is not.
The soak interval needs to be long enough for the surface to stop ponding — typically fifteen to thirty minutes — and short enough that the programme still finishes before the sun is high. Twenty minutes is a reasonable default on most soils.
Choosing the window
The clock matters almost as much as the minutes. Water applied at midday loses a substantial share to evaporation before it reaches the soil, and on a hot day with a breeze that share can be a third. Early morning is the best window: wind is usually at its lowest, temperature and evaporation are low, and anything that lands on foliage dries quickly, which reduces the time leaves stay wet and the disease pressure that comes with it.
Evening watering is the second choice and it is not a bad one, but it leaves foliage wet overnight, which is exactly the condition that encourages mildew and black spot on susceptible plants. If you water in the evening, aim at the soil rather than the leaves — which in practice means a drip line or a soaker hose rather than a sprinkler.
| Planting | Depth per session | How often | Note |
|---|---|---|---|
| Established lawn | 12–15 mm | Once or twice a week | Deep and infrequent beats light and daily |
| Newly sown lawn | 5 mm | Daily until established | Shallow, because the roots are at the surface |
| Vegetable bed | 15–20 mm | Twice weekly when dry | More in fruiting, less at either end |
| Established shrubs | 20–25 mm | Weekly when dry | Deep watering, then leave it alone |
| Containers | To run-off | Daily in hot weather | Volume matters less than frequency |
| Newly planted trees | 20–30 mm | Weekly for the first season | At the root ball, not the trunk |
A worked example, including the split
- Area and depth to litresSeventy-six square metres at twelve millimetres is 912 litres — the multiplication is area times depth and nothing else.
- Correct for efficiencyAt 75% efficiency the system has to deliver 1,216 litres for 912 to land in the root zone.
- Convert to minutesAt 12 litres a minute that is 101 minutes of running time.
- Check it against the soilFifteen minutes is the most this ground takes in one go, so 101 minutes becomes seven cycles of about fourteen and a half minutes each.
- Put it on the clockSix twenty-minute soaks between them take the programme to 221 minutes — three hours and forty minutes from a six o'clock start, finishing before ten.
That last line is the one people do not expect. A watering session that needs a hundred minutes of running occupies nearly four hours of the morning, and if it starts at six it is finished by ten, which is fine. Starting the same programme at four in the afternoon puts the last cycle in full sun, where a quarter of it evaporates on the way down.
What to measure, and what to ignore
Only three of these numbers need measuring properly, and the flow rate is the one that is always guessed. Put a bucket under the sprinkler, run it for a minute, and measure what came out — or time how long it takes to fill a container of known volume. Flow rate varies with pressure, which varies with how many other taps are open and how long the hose is, so measure it in the configuration you will actually use.
Area comes next, and it is the area the sprinkler covers rather than the whole plot: an oscillating sprinkler throwing a rectangle into a circular bed is watering the rectangle. The lawn area calculator handles the usual shapes and lets you subtract the beds and obstacles inside them, and the garden area calculator does the same where the ground is being planted rather than grassed.
If the water is going through a drip system rather than a sprinkler, the drip line length figure is what determines the flow rate to enter here, because the emitters are rated per metre. Mulch changes the whole picture by slowing evaporation off the surface, and the mulch volume calculator is the cheapest way to reduce the number you are about to compute. Where the beds are raised rather than at grade, the raised bed soil figure tells you how much of the applied depth the bed can actually hold.
Do
- Measure the flow rate with a bucket and a clock, in the configuration you will use.
- Water early morning, when wind and evaporation are lowest.
- Split any run longer than the soil can absorb into cycles with soak gaps.
- Water deeply and less often, to push roots down rather than keep them at the surface.
- Reduce or skip a session after decent rainfall.
Avoid
- Guessing the flow rate, which is the single largest source of error here.
- Running one long session on a soil that cannot take it, and watching the surplus run off.
- Watering at midday, where evaporation takes a large share before it lands.
- Light daily sprinkling, which keeps roots in the top few centimetres where they dry out first.
- Watering to a schedule through a wet week rather than checking the soil.
Where the numbers stop being true
This assumes uniform coverage — that every part of the area receives the same share. Real sprinklers do not achieve that: they overlap in the middle and under-deliver at the edges, and a wind pushing the pattern to one side can double the application on one half while halving it on the other. Catching this requires catch cans, which is a worthwhile hour with a set of straight-sided tins if you are watering anything valuable.
The efficiency figures are typical rather than measured, and the infiltration limit is one you have to supply from observation rather than from a table, because it depends on compaction, organic matter and how dry the ground is to begin with. Bone-dry soil actually accepts water more slowly at first, not faster, until the surface has wetted.
Rainfall is not modelled at all. A session after fifteen millimetres of rain is pure waste, and the honest way to handle it is a rain gauge or a water-balance check rather than a calendar. Finally, this works on one session: a programme with several zones running in sequence on one controller needs the same arithmetic per zone, added together, and checked against the total flow the supply can deliver.
How the figures are produced
Area multiplied by depth gives litres directly. That figure is divided by the efficiency to give the volume that has to be delivered, and divided again by the flow rate to give minutes. If those minutes exceed the longest single run the soil will accept, they are split into equal cycles with a fixed soak interval between them; the cycles are as few as possible while staying under the limit, and each is the same length so the programme is repeatable on a simple controller.
Questions people ask
Why is the total so much longer than the running time?
Because of the soak gaps. A hundred minutes of running split into seven cycles carries six twenty-minute gaps, which is two hours of the programme doing nothing. That is the point: the gaps let each application soak in before the next arrives.
Can I just run it in one go and accept the run-off?
You can, and on sandy soil with a low application rate it is fine. On clay, on a slope, or with a sprinkler that delivers faster than the ground absorbs, most of the water beyond the first fifteen minutes is going somewhere you did not intend.
How long should the soak be?
Long enough for the surface to stop ponding — fifteen to thirty minutes on most soils. Shorter and the next cycle lands on ground that is already saturated; much longer and you run out of morning.
Does a drip system need cycling?
Much less often, because the application rate is low enough that infiltration keeps up. Enter a high efficiency and a long maximum run and the calculation will usually return a single cycle, which is the honest answer for drip.
Should I water every day?
No. Deep and infrequent builds roots that follow the water down; light and daily keeps them in the top few centimetres, which is the layer that dries out first and the reason lawns watered daily still brown in a heatwave.
What about a hosepipe ban?
A watering ban almost always exempts newly planted material and sometimes exempts drip systems and soaker hoses, because they deliver water without waste. If you are under restrictions, the efficiency figure is the thing to improve rather than the duration.
My pressure drops when I run two zones. What then?
The flow rate drops, so the run time lengthens, and the sprinkler pattern degrades as well. Measure the flow with both zones running if that is how you will use it, and expect the coverage to get worse in a way no calculation can correct for.