Does companion planting actually work?

Some of it works and most named pairings do not. Five mechanisms hold up: physical support, shade, nitrogen fixing, attracting beneficial insects, and trap cropping. The specific folk pairings — basil improving tomato flavour, one herb repelling another crop's pests — have no mechanism and no evidence. The distinction is worth knowing because it changes what you do.

The mechanisms that genuinely exist

When companion planting works, it is usually one of these five things.

Physical support. Beans climb corn. Peas climb anything. Not subtle and not disputed.

Shade and living mulch. A sprawling squash shades the soil at the base of taller crops, keeping it cooler and damper and suppressing weeds. This is the real mechanism behind the traditional corn, beans and squash planting.

Nitrogen fixing. Beans and peas host bacteria that convert atmospheric nitrogen into a usable form. The effect on a neighbour in the same season is modest — most of the nitrogen stays in the legume until it is turned in — but over a rotation it is real. It is also why our bush bean guide says not to feed them nitrogen.

Attracting beneficial insects. Small open flowers — dill, fennel, cilantro when it bolts, alyssum, yarrow — bring in hoverflies, lacewings and parasitic wasps, whose larvae eat aphids. Probably the most useful item on this list for a vegetable garden.

Trap cropping. A plant a pest prefers, drawing it away from the crop. Nasturtiums for blackfly is the classic. It works — with the caveat that you now have a plant covered in blackfly and have to deal with it.

What the evidence is thinner on

Repelling pests by smell. The idea that strongly scented herbs mask or repel pests is appealing and mostly unsupported at garden scale. Disrupting an insect's search pattern with dense mixed planting has some support; a single basil plant repelling anything does not.

Improving flavour. No mechanism, no evidence.

Long lists of specific antagonisms. Most published "never plant X with Y" tables trace back to a small number of early-twentieth-century books rather than to trials. Mississippi State and Arizona extension services both treat the genre sceptically, while Fine Gardening and others defend it — the disagreement is genuine and worth knowing about before you take any chart as settled.

What should not be planted next to tomatoes

Three things, and only one of them is about chemistry.

Potatoes. Same family, and they share late blight. This is the one that matters. It is a rotation rule as much as a neighbour rule: do not grow tomatoes where potatoes were last year either. Our potato guide and tomato guide both carry it.

Fennel. Genuinely inhibits the growth of many neighbouring plants. Grow it in its own corner.

Sweet corn, on the sunny side. Corn at 10-inch spacing grows tall enough to shade a tomato bed from midsummer. This is geometry rather than chemistry, and it is the most common real "bad companion" in a garden.

Everything else on the usual tomato avoid-lists — brassicas, dill, walnut aside — is weakly supported at best.

The avoid-list that does hold up

Same family, same bed, year after year. Tomatoes and potatoes share blight. Brassicas share clubroot. Onion-family crops share white rot. This is rotation, and it matters a great deal more than any pairing.

Fennel, as above.

Tall next to short, on the sunny side. In the northern hemisphere, corn or staked tomatoes on the south side of a bed shade everything behind them from midsummer. How to plan a vegetable garden layout covers getting this right.

Heavy feeders together. Two hungry crops in the same square foot compete. Brassicas and corn in adjacent squares will both disappoint.

What to do instead of following a chart

The things that reliably improve a bed are duller than companion planting and work better.

Get the spacing right. Crowded plants compete for light, water and nutrients, and poor airflow drives foliar disease. The plant spacing calculator works from published spacings rather than guesses.

Rotate by family. Four groups, one bed each, moving along one per year.

Grow flowers among the vegetables. Not for repelling anything — for the hoverflies and parasitic wasps. Let some cilantro and dill bolt rather than pulling them.

Mix crops within a bed. Trials consistently show lower pest pressure in mixed plantings than in monocultures. That is a diversity effect, and you get most of it by growing several unrelated crops together without needing any chart to tell you which.

Feed and water according to the crop. A well-grown plant resists pests far better than a stressed one beside the right neighbour.

How our crop pages handle this

Each crop guide lists traditional companions and crops to keep apart, because readers expect them and some of the pairings are useful. Every one of those pages also says plainly that companion planting pairings are garden tradition rather than settled science, and that spacing, sun and water do far more for a crop than its neighbours do.

That is the honest position: the lists are a starting point for arranging a bed, not a system for managing pests.