What plant spacing actually decides
Spacing is a yield decision disguised as a layout decision. Plant closer than the crop wants and every plant competes for the same light, water and nitrogen, so you get more plants and smaller ones; plant wider and each performs better while the bed produces less. The spacing on a seed packet is the point the breeder found the two effects balance for a full-size harvest.
Square foot gardening, popularised by Mel Bartholomew, made that number easy to use. Instead of rows with paths between them, you divide the bed into one-foot squares and put a whole number of plants in each: 16 at 3 in spacing, 9 at 4 in, 4 at 6 in, 1 at 12 in. The reason those numbers are perfect squares is simply that 12 ÷ 3 = 4 and 4² = 16 — the grid is two-dimensional, so halving the spacing quadruples the count.
That is also why spacing is such a violent lever. Going from 6 in to 4 in spacing is a one-third reduction in distance and a 125% increase in plant count. Reading the plants-per-square-foot figure at the top of the results before you commit to a spacing is the fastest way to see whether a bed will hold the crop you have in mind.
Grid versus offset, and why the textbook gain does not always appear
The square grid puts plants at the corners of squares of side s. Each plant claims an area of s², so density is 1 ÷ s² per square inch, or (12 ÷ s)² per square foot.
The offset or triangular layout shifts every second row sideways by half a spacing, which lets the rows sit closer together while keeping every plant the same distance from its neighbours. The row pitch drops to s × √3 ÷ 2 = 0.866s, so each plant claims 0.866s² instead of s². That is the famous 15.5% density gain, since 1 ÷ 0.866 = 1.1547.
The gain is asymptotic, and small beds often do not get it. It only materialises if the closer row pitch lets an extra row fit inside the bed, because the offset rows themselves hold one fewer plant than the full rows. In a 4 by 8 ft bed at 12 in spacing the grid puts down 4 rows of 8 = 32, while the offset pattern still only fits 4 rows — 48 ÷ 10.39 = 4.6 — and those alternate 8, 7, 8, 7 for a total of 30. The offset layout is worse here. At 3 in spacing in the same bed it fits 18 rows against the grid's 16, and comes out at 567 against 512, which is the 10.7% gain finally showing up.
The comparison table in the results computes both counts for your actual rectangle at seven common spacings, which is the only reliable way to decide. Do not assume the offset pattern wins; check the row.
Seeds are the plant count run backwards through two losses. Multiply by one plus your thinning allowance, then divide by the germination rate, then round up. At 32 plants, a 20% thinning allowance and 85% germination that is 32 × 1.20 ÷ 0.85 = 45.2, so 46 seeds. Note the division: a low germination rate raises the seed count far faster than a modest thinning allowance does.
Worked example: bush beans in a 4 by 8 ft raised bed
The bed is 96 in by 48 in inside the frame. The packet says thin bush beans to 4 in apart. Germination on the packet is 85%, and you want to sow 20% extra so you can thin to the strongest seedling.
- Bed area. 96 × 48 ÷ 144 = 32 square feet.
- Plants per square foot. floor(12 ÷ 4) = 3, and 3² = 9 per square foot.
- Columns. floor(96 ÷ 4) = 24.
- Rows. floor(48 ÷ 4) = 12.
- Plants in the bed. 24 × 12 = 288, which is 32 sq ft × 9 — the two routes agree, as they must.
- Offset check. Row pitch = 4 × 0.866 = 3.46 in, so 1 + floor(44 ÷ 3.46) = 13 rows, alternating 24 and 23 plants: 7 × 24 + 6 × 23 = 306 plants, 18 more than the grid.
- Seeds. 288 × 1.20 ÷ 0.85 = 406.6, so buy 407 seeds — and at the offset count, 306 × 1.20 ÷ 0.85 = 432.
At 4 in spacing the offset layout does pay, unlike the 12 in case. It gains 18 plants, which is 6.25% rather than the theoretical 15.5%, because the bed is only 12 spacings wide and the boundary effect eats most of the advantage.
Reading the result and sanity-checking the spacing
Check the plants-per-square-foot figure against the crop first. If it says 16 and you are planting tomatoes, the spacing input is wrong — you have entered the in-row seed spacing rather than the final thinned distance. Packets frequently give both, as "sow 1 in apart, thin to 4 in", and the number this calculator wants is the second one.
Then look at whether the whole-bed count is realistic to manage. Three hundred bean plants in 32 square feet is a real harvest and a real amount of picking; 512 radishes at 3 in spacing mature within about a week of each other, which is why radishes are usually sown in successions rather than filling a bed.
Where the spacing exceeds 12 in the plants-per-square-foot figure becomes a fraction, and the square foot gardening convention flips: instead of several plants per square you get one plant per several squares. An 18 in crop is 0.444 per square foot, or one plant per 2.25 squares, which in practice means one plant per 4 squares on the grid because you cannot plant a quarter of a square.
Finally, treat the square-feet-for-a-target output as a bed-sizing tool rather than a planting instruction. It divides your target count by the density this bed actually achieves, so it already includes the edge losses of a bed this shape. A long narrow bed and a square one of the same area will give slightly different answers, and that difference is real.
Square foot gardening spacings for common crops
| Spacing | Per square foot | Per 4 × 8 ft bed | Typical crops |
|---|---|---|---|
| 3 in | 16 | 512 | Radish, carrot, spring onion, mesclun |
| 4 in | 9 | 288 | Beetroot, spinach, bush bean, turnip |
| 6 in | 4 | 128 | Lettuce, chard, garlic, bok choy |
| 12 in | 1 | 32 | Tomato, pepper, cabbage, broccoli |
| 18 in | 0.44 | 10 | Courgette, aubergine, indeterminate tomato |
| 24 in | 0.25 | 8 | Winter squash, melon, rhubarb |
The four classic densities are from the square foot gardening method; the two widest rows follow the same formula but are planted one per several squares. Bed counts are floor(96/s) × floor(48/s).
Where a spacing calculation goes wrong
- Using sowing spacing instead of thinned spacing. The two differ by a factor of three or four on root crops, which is a factor of nine to sixteen in plant count.
- Measuring the frame instead of the soil. A bed built from 2 in stock loses 4 in on each dimension. On a 4 ft bed that is one whole 4 in row.
- Ignoring what the plant does above ground. Spacing controls root competition, but a caged indeterminate tomato at 12 in centres shades everything behind it. Put tall crops on the north edge in the northern hemisphere.
- Treating trellised crops as if they need ground area. Cucumbers and pole beans grown vertically can be spaced far tighter than the packet's ground-culture figure, because the packet assumes they sprawl.
- Forgetting succession. A bed filled to capacity on day one produces one harvest. Splitting it into thirds sown three weeks apart produces the same total over six weeks, which is usually more useful.
- Trusting old seed at the printed germination rate. Rates on the packet are from testing at packing. Two- and three-year-old seed of alliums and parsnip in particular falls off sharply, and the seed count is divided by that number.
Where the grid method fits among other approaches
The square grid is a bed method, not a field method. It works because a raised bed is never walked on, so soil structure survives and roots can occupy the whole volume; that is what lets you drop the paths that conventional row cropping needs. In a field, where a tractor or a hoe must pass, wide rows with tight in-row spacing win on total yield per hour of labour even though they lose on yield per square foot.
The offset triangular pattern is what commercial growers use inside a bed for exactly the density reason above, and it is also how orchard and paver layouts are set out. If you are laying anything else on a grid or an offset pattern in the same yard, the geometry carries over directly — the baseboard and trim linear feet calculator handles the linear version of the same edge-margin problem indoors.
The other numbers a bed needs are water and heat. Irrigation run time depends on bed area rather than plant count, and the same bed that holds 288 bean plants and 32 square feet of soil is the one whose watering you are sizing. If the yard also has a pool, the pool heating cost calculator covers the other big seasonal running cost, and for the hours you spend out there in summer the UV index sunburn time calculator is the relevant safety check. Households planning for dry spells should also look at the emergency water storage calculator, since a productive bed is a standing water commitment.
Terms on the seed packet
- In-row spacing
- The final distance between plants along a row, after thinning. This is the number to enter here for an intensive bed.
- Row spacing
- The distance between rows in conventional culture, which usually includes access room for a hoe. Intensive beds discard it and use the in-row figure in both directions.
- Station sowing
- Dropping two or three seeds at each final plant position and thinning to the best one, rather than sowing a continuous line. It is why the thinning allowance exists.
- Germination rate
- The share of seeds that produce a viable seedling under test conditions, printed on commercial packets. Field results are normally lower, especially in cold or dry soil.