The first thing I tell anyone planning raised beds in a small yard is that the beds are not the hard part. The paths are. Get the walking space wrong and you end up with a set of boxes you cannot comfortably reach into. That defeats the whole point of raising them.
Everything below assumes space is genuinely tight. If you have an acre, the constraints are different. Most of this will not apply.
The real constraint is reach, not area
A raised bed exists to solve three problems: bad ground, bad posture, and bad drainage. In a small yard it also has to solve a fourth, which is that you cannot afford a bed you have to walk around to use.
So start from the body, not the plot. An adult can reach roughly 60 cm into a bed from one side without straining. Work from that. It gives a comfortable maximum bed width of about 1.2 m when the bed is accessible from both sides, and about 60 cm against a wall or fence.
Most beginners build beds twice that wide. They look generous on a plan. They are miserable in August.
Paths matter just as much. A main path of 90 cm lets two people pass and takes a wheelbarrow. Secondary paths of 45 to 60 cm are fine for foot traffic. Anything narrower than 45 cm becomes a trip hazard once plants flop over the edge.
Depth: the number that decides what you can grow
Raised bed depth is usually quoted as one figure. That is unhelpful. The right depth depends entirely on what is going in it.
Leafy crops, herbs, and most salad have shallow roots. They will manage in 15 to 20 cm of soil, provided the ground beneath is not concrete. Tomatoes, peppers, and bush beans want 25 to 30 cm. Root crops are the demanding ones. Carrots and parsnips need 30 cm of loose, stone-free soil or they fork. Maincrop potatoes want more than that.
For a general-purpose bed in a small garden I would build to 30 cm of soil depth. It covers almost everything. It drains well. And it does not require a soil delivery you cannot afford.
If the bed sits on soil rather than paving, build shallower. Let the roots go down into the ground below. Clemson’s guide to raised beds covers the open-bottom and lined variants and when each is appropriate.
Materials, and what each one actually costs you
Timber is the default and remains a reasonable choice. Untreated oak or larch lasts perhaps five to ten years. Treated timber lasts longer. The food-safety worry that circulates about it concerns older preservatives no longer sold for garden use. Modern pressure-treated timber is a different formulation. If you are uneasy, line the inside with heavy polythene. Or accept a shorter lifespan and use untreated boards.
Do not use old railway sleepers or reclaimed industrial timber unless you know what it was treated with. That is the one material risk I will not take on a bed growing food.
Galvanised metal beds are the pragmatic choice for a small yard. They are thin, so they waste no space on board thickness. They last decades. And they are usually sold in modular panels that assemble with bolts. The trade-off is heat. Metal conducts, so the soil warms faster in spring and dries faster in summer. In a hot climate that means more watering.
Concrete block is the cheapest per metre if you can collect it, and it holds moisture well because of its thermal mass. It is also heavy, and it is essentially permanent. Once laid, that bed is not moving.
Fabric beds are cheap, light, and fold away, which makes them useful for a temporary layout. They last a few seasons at most and they dry out quickly at the edges.
Working out how much soil you need
This is the calculation people get wrong. It is the one that costs real money.
Multiply length by width by depth, all in metres. A bed 2.4 m long, 1.2 m wide, and 0.3 m deep is 0.864 cubic metres. Round up, because soil settles.
That is roughly 860 litres. About 30 bags of the size sold at a garden centre. Once you see that number, buying in bulk by the cubic metre stops looking expensive.
Do not fill a bed entirely with compost or bagged potting mix. Compost is a soil amendment, not a growing medium on its own. It shrinks as it decomposes. It holds too much water. And plants in pure compost tend to grow soft and floppy. A common starting mix is roughly 60 per cent topsoil, 30 per cent compost, and 10 per cent coarse material such as grit or bark for drainage. The RHS covers this in its guide to soil types.
If you are using topsoil you bought or had delivered, test it. A basic pH and nutrient test costs little. It tells you whether the soil you just paid for is any good. Clemson’s soil testing factsheet explains what the results mean and how to take a representative sample.
Layouts that earn their footprint
In a genuinely small yard, the classic layout is a border arrangement. Beds run along the perimeter with a single central path. Not rows of beds with paths on all sides. You sacrifice access from one side, so keep those beds to the 60 cm width.
Table beds on legs are worth considering if bending is the limiting factor rather than space. The trade-off is soil volume. A 1 m by 0.5 m bed on legs holds a fraction of what a ground-level bed of the same footprint holds. It suits herbs and salad. Not tomatoes.
Vertical structures work for climbers and trailing crops. They work badly for almost everything else. A wall-mounted planter holding lettuce will cook it in full sun. A wall in direct light is one of the hottest surfaces in a garden. Put vertical planters on a north- or east-facing wall. Keep the sun for the ground beds.
Avoid tiered stands in a small yard unless you can still reach the back tier. The rear row of a three-tier stand is usually the one that fails.
What to grow, honestly
Raised beds suit crops that like warm, free-draining soil. Lettuce, chard, radish, bush beans, peppers, and herbs all do well in 25 cm. I grow most of those every year.
They suit root crops only if you built deep enough. If you did not, grow them elsewhere. Forked carrots are not worth the space.
Large, hungry, vining crops are a poor fit in a small bed. Courgettes and pumpkins sprawl well beyond their allotted space. They will smother whatever is next to them. If you want them, give them their own bed at the edge of the plot and train them outward.
Tall crops cast shade. A bed of sweetcorn or a teepee of climbing beans on the south side of a small garden will shade the rest of it for most of the day. Put the tall things on the north side. I learned this one after losing a bed of salad to a bean teepee.
The mistakes I see most often
Beds built too wide to reach across. It is the most common error by a wide margin. And it is unfixable once the soil is in.
Filling with pure compost, then wondering why everything grew soft and the level dropped by a third in a season. I have watched this happen.
No drainage consideration on a bed sitting over paving. Water needs somewhere to go. If there is no ground beneath, the base needs a permeable layer. The bed also needs to sit slightly off the surface.
Overcrowding. Seedlings look sparse in a big bed in April. The temptation to double up is strong. By July the air cannot move between the plants. That is where fungal disease starts.
Ignoring the sun path. A bed placed in the only spot available, rather than the sunniest spot, will underperform all season. Track the sun across the site for a day before you commit. That is the cheapest survey you will ever do.
Maintenance, briefly
Top up the soil level each spring. It settles, and the level drops. Two to three centimetres of compost as a mulch does that job and feeds the soil at the same time. Clemson’s composting factsheet covers what is safe to add.
Check the boards or panels annually. A rotting board replaced in autumn costs a few pounds. The same board ignored for a year costs you the soil behind it. I check mine every October.
Refresh the soil every few years. Do not replace it. Beds improve with age if you keep adding organic matter. A bed older than five years is usually better than a new one.