A vegetable plot is not a rectangle you fill; it is beds you can reach into, separated by paths you can work from. Give this the outside dimensions of your plot and it works out how many beds fit, how many cropping rows run down each one, and — the figure nobody quotes but everybody needs — how many metres of row you will actually be sowing, weeding and harvesting this season.
The plot
Set by how far you can comfortably reach from the path.
Result
The number nobody quotes
Row length is the figure that predicts how much work a plot actually is. Area tells you how much ground you have; row length tells you how much bending, hoeing, watering and picking you have signed up for, because all of it happens along a row rather than per square metre. A 12 m × 10 m plot laid out badly can have 150 m of row in it. Laid out well it can have nearly 300 m in the same space, and twice the harvest for roughly the same effort per metre.
Bed width comes from your reach, not your plot
This is where most layouts go wrong. A 2 m bed looks efficient on paper because it wastes less space on paths, and then you discover in June that the middle of it cannot be reached from either side without standing on the soil you have just dug. The correct width is set by the distance from the path edge to your fingertips when crouching, doubled.
A path is not lost ground. It is where you stand to work, where the barrow runs, where water drains and where the soil is not compacted by your feet. Every attempt to reclaim path width shows up later as a bed you cannot reach the middle of, or soil you have walked into a crust.
Narrower paths with wider beds also means fewer positions to rotate crops between, which matters more than the extra 2 m² of growing space.
How wide a path really needs to be
Path width is decided by what has to travel along it and how often you will be kneeling on the bed beside it. There is no single answer, but there is a clear wrong one: anything you cannot put a boot on sideways.
| Path needs to take | Minimum width | Why |
|---|---|---|
| Kneeling and reaching only | 30 cm | Enough to put both knees on the bed edge and lean in |
| A wheelbarrow, occasionally | 45 cm | Barrow wheels are usually 30–40 cm apart |
| A barrow and a person passing | 60 cm | Comfortable rather than a squeeze |
| Wheelchair or mobility aid | 90–120 cm | Turning circle matters more than straight running |
| Between permanent raised beds | 60–90 cm | Secure edging means the path takes all the traffic |
Rows per bed
Once the bed width is fixed, the crop's row spacing decides how many rows fit. The calculator keeps 15 cm off the bed edge before it starts counting, because the outermost plants sit closer to dry soil and competition from whatever is on the other side of the path.
This is also where the row-spacing and plant-spacing calculations part company. Rows are about access and airflow; the distance along the row is about plant size. Only one of them changes when you switch from carrots to cabbages, and mixing the two up is why a bed planned for salad leaves looks absurd when planted with brassicas.
Orientation: which way should rows run
On a flat plot with nothing shading it, run rows the length of the plot to minimise turning at the ends. Everything else beats that rule: on a slope, run rows across the slope rather than down it, because down-slope rows become channels the first time it rains properly. Where tall crops are involved, run rows north to south so they shade their neighbours least.
Do
- Run rows across a slope so they hold water rather than shed it.
- Align rows north to south where anything tall is growing.
- Put the tallest crop on the side that will not shade the rest.
- Keep perennial crops to one bed so annual digging never disturbs their roots.
Avoid
- Running rows down a slope because it matches the shape of the plot.
- Letting a row of sweetcorn or Jerusalem artichokes cast a day-long shadow across its neighbours.
- Mixing perennial and annual crops in the same bed and calling it rotation.
- Turning the layout every season, so nothing ever settles into a routine you can maintain.
A worked example with a rotation behind it
Here is a plot planned not for one season but for four beds carrying a rotation, which is what actually decides how many beds you need.
- Start with the plot10 m by 12 m: 120 m², which sounds enormous until you subtract the paths.
- Fix the bed width from reachBeds of 1.2 m with 0.4 m paths. Together each unit is 1.6 m, so (10 + 0.4) ÷ 1.6 = 6 beds.
- Decide rows per bed for the main cropBrassicas at 45 cm: (1.2 − 0.15) ÷ 0.45 = 2.33, so 3 rows per bed.
- Add up the row length6 × 3 × 12 = 216 m of row. Six beds also divides neatly: two for brassicas, two for roots, one for legumes, one for everything else, rotated annually.
- Check the tradeGrowing area is 6 × 1.2 × 12 = 86.4 m², which is 72% of the plot. The missing 28% is the price of being able to reach all of it without standing on it.
What the layout does to the season
Row length is also a planning figure. Seed is sold by weight and recommended per metre of row as often as per square metre, so once you know you have 216 m of row you can buy seed accurately rather than hopefully. The seed quantity calculator handles that conversion, and the plant spacing calculator takes over when it is plugs rather than seed going in.
Before you mark it out
- Walk the plot after rain and note where water stands or runs; that decides orientation.
- Mark out beds with string before digging anything permanent, and live with it for a week.
- Decide now which beds are perennial, because they should never be dug again.
- Put the paths where you actually walk rather than where the plan looks tidy.
- Check the plot area against the garden area calculator if the boundary is anything other than four straight sides.
Where the numbers stop being true
It assumes straight beds of equal width on level ground, which is the layout most people should start with and few plots actually are. Once the ground slopes, or the plot tapers, or there is a tree in it, treat the answer as an upper bound and mark out fewer beds than the arithmetic allows.
It also assumes every row is a full row. In practice one bed will be part-shaded and given over to something that does not mind, one will be half-full of a crop that failed, and one will have a path cut across it because that is where you kept walking. Which is fine: the value of the calculation is that you found out you had 288 m of row to look after before you sowed it, rather than halfway through July.
How the arithmetic works
Beds across: (plot width + path width) ÷ (bed width + path width), rounded down. That identity works because n beds have n − 1 internal paths, so adding one path width to the plot converts the problem into clean units. Rows per bed: (bed width − 0.15 m) ÷ row spacing, rounded down, plus one. Total row length: beds × rows per bed × bed length. Growing area: beds × bed width × bed length, and the percentage divides that by the plot area.
Six beds 1.2 m wide with 0.4 m paths occupy 6 × 1.2 + 5 × 0.4 = 9.2 m, not 6 × 1.6 = 9.6 m. Adding one path width to the plot width makes the division exact: (10 + 0.4) ÷ 1.6 = 6.5, which floors to six and leaves the leftover 0.8 m to absorb into the path on one side.
Questions people ask
Can I make the beds wider and the paths narrower?
You can go to about 1.4 m before reaching the middle becomes uncomfortable, and down to 30 cm on paths if nothing with wheels needs to pass. Beyond that you trade reach for space and end up compacting the soil you were trying to gain.
Do I need paths at all if I grow in raised beds?
Yes, unless the plot is narrower than about 1.2 m in total. Even a bed you never step into needs somewhere to stand to work on it, and the soil between two raised beds takes all the traffic for both of them.
Should I dig the beds or use no-dig?
Either. The layout maths is identical. No-dig simply means paths are more permanent, so get the dimensions right first time, because moving a mulched path is more work than moving a dug one.
How many crops should rotate through the beds?
Four is the traditional answer — roots, legumes, brassicas and everything else — because that gives each family three years away from its own pests. With fewer than four beds you rotate on a shorter cycle, which still helps.
What do I do with the leftover strip?
Whatever it is. Because the division rarely comes out exact, most plots end with a narrow remainder. Use it for herbs, a perennial edge, or widen one path — leaving it as an unusable sliver of bare soil is the one genuinely wrong answer.