Building a Dry Stone Terrace by Hand: What Holds and What Fails

In September 2023, after three months without a drop of rain, the first serious storm brought a hundred and twenty millimetres in two days and two of my terraces went with it. Building a dry stone terrace is slow work with a simple method, and the walls that held were the ones I had taken time over — the ones that failed were built in a hurry in 2021, when I was racing the rains. The difference was entirely in the foundation.

What this covers

• Why unmortared stone outlasts a rigid wall on a slope that moves
• The foundation trench, and the backward tilt that turns soil pressure into compression
• The one-over-two rule, and the failure it prevents
• What I learned when two of mine washed out
• A behaviour of dry soil I had not anticipated

Why dry stone and not mortar

A mortared wall is rigid. On a slope that shifts a little every winter with frost and heavy rain, a rigid wall cracks and then fails all at once.

A dry stone wall flexes. Each stone can move a few millimetres against its neighbours without the structure losing integrity. And it drains instantly — there is no mortar to trap water behind it.

The walls on this land that nobody living built are still standing. That is the argument, and it is a better one than anything I could construct.

The foundation

This is where it is decided, and it is where mine failed.

I dig a shallow trench along the contour, below the existing soil surface, down to something that does not move — compacted subsoil or rock. Then I set the largest and flattest stones I have directly into it, each tilted very slightly back into the hillside.

That backward tilt is the whole principle. It turns the weight of the soil behind the wall into a force pressing stones together, instead of one pushing them out.

The two walls I lost in 2023 failed here. I had built them quickly in the autumn of 2021, ahead of the rains, and I had not dug deep enough to reach stable ground. They sat on loose material. When the water came, the loose material moved and everything above it followed.

The lesson is unglamorous: better no terrace than one that will go, because when it goes it takes the soil with it. I now do not start a wall unless I have the time to do the foundation properly.

Building up

Every stone above the foundation sits across the joint between two stones below it — the same principle as brick bonding.

A continuous vertical seam is the commonest reason these walls fail. It creates a straight fracture line the whole wall can split along.

Smaller stones and rubble go behind the face stones as packing, not in front. They fill voids and lock the structure. Rubble showing on the face is a wall that will shed stones.

What a terrace actually does here

Beyond making flat ground, its work on a dry slope is to slow water.

I did not fully appreciate how much this matters until I compared two pieces of ground through a winter. On cleared, bare ground, a heavy rain on a slope runs — it cuts small channels, carries silt, pools at the bottom. Where there was vegetation and a wall to hold it, the water goes in.

On shallow limestone soil on a slope, with 750 millimetres arriving almost entirely between November and March, that is the most consequential thing that happens on my land all year. Soil that leaves the hillside does not come back, and water that runs off instead of soaking in is water I will not have in July.

There is a second effect I only mapped properly after the 2023 drought. Going through my food forest tree by tree that September, the pattern was unmissable: trees at the foot of terraces survived almost without exception, trees on exposed ridges died almost without exception. Two trees of the same species and age, fifteen metres apart, one alive and one dead.

Water on this ground does not distribute. It concentrates in specific places, and terraces create those places deliberately.

A wall is not the only way to do that. On gentler ground, or where I have no stone to spare, swales and berms achieve the same thing with a spade and considerably less of my winter — the principle is identical, only the material changes.

The thing I had not anticipated

The failures did not happen in July. They happened in September, with the first heavy rain after the long drought.

Soil that has been without water for three months becomes water-repellent at the surface. It does not absorb. That first serious rain ran over the ground instead of into it, and the volume of surface water was far greater than anything I had designed for.

I have not seen this discussed in the guides I have read, and it is worth knowing if you build in a climate with a long dry season: your walls will be tested hardest not by the wettest month, but by the first storm after the driest one.

What I keep from this

• The foundation decides everything; if you cannot do it properly, do not start
• Dig down to something that does not move, not just below the topsoil
• Tilt every foundation stone back into the hill
• One stone over two, always; a vertical seam is a fracture line
• Packing goes behind the face, never in front
• The real test comes with the first storm after the dry season, not in midwinter

Questions people ask

How long does a dry stone terrace last? The oldest ones on this land were built by people nobody now remembers and they are still holding. Mine are too young to say. The two that failed were two years old and badly founded, which tells you where the risk sits.

Do I need special stone? I use what comes out of the ground here, which is limestone in every awkward shape. Flat stones are worth saving for the foundation course and the top. Beyond that you work with what you have and spend the time finding which stone fits which gap.

How much can one person build? Slowly. I have built all of mine alone, in winter, in stretches of a few metres. It is not work that rewards hurry, and the two I hurried are the two I rebuilt.

Does it need drainage behind it? Not in the way a mortared wall does. The whole face drains. What it does need is packing behind the face stones so water moving through does not carry fine material out with it.

Related reading

• What the Jackals Taught Me About Water — where water hides on a hillside, and how the old walls found it first
• Dry Farming Vegetables Without Irrigation — what the stored moisture behind a terrace is for
• Rainwater Harvesting with Swales and Berms — the earthwork version of the same principle, for ground that does not need a wall

Before you go

I send one letter a month from this hillside — what was tried, what failed, and what the season is doing. No schedules, no promises about yields.