The Pye Garden
Practical Guides for Home Gardeners, Houseplants & Organic Cultivation
Home / Vegetable Gardening • PLANT CARE GUIDE

Companion Planting: Separating Useful Pairings From Folklore

The Pye Garden Editorial Team • November 20, 2026
Practical Guide
Companion Planting: Separating Useful Pairings From Folklore
Cultivation guide: Companion Planting: Separating Useful Pairings From Folklore.

The chart on the back of a seed packet manages to pair a dozen vegetables with tomatoes and give a reason for none of them. Basil, marigold, parsley, carrot, chive, nasturtium — the arrows are confident, and the effect in a real bed is usually invisible. A few of those pairings rest on something measurable. Others are simply two plants that tolerate each other, which is a weaker claim wearing the same clothes.

What good pairings actually do

Five mechanisms do most of the genuine work, and each one predicts when a pairing helps and when it will not.

Pest confusion. Insects locate hosts by smell and silhouette. The carrot fly follows the scent of carrot foliage, and a dense row of onions or leeks nearby dilutes that cue. It reduces attack rather than eliminating it, and it works best in a small garden where the masking scent is a large fraction of the air the fly samples.

Habitat for predators. Dill, fennel, cilantro, and parsley left to flower, plus sweet alyssum and buckwheat, supply nectar and pollen to hoverflies, lacewings, and the small parasitic wasps that hunt aphids and caterpillars. The adults feed on flowers while their larvae do the killing, so the flowers have to be open and within a few meters of the pest. A single dill plant at the far corner of the lot does little.

Shade. A trellised cucumber or a block of corn keeps midsummer lettuce a few degrees cooler and delays bolting by a week or two. The same canopy starves the lettuce of light in September. Shade is a seasonal tool, not a permanent arrangement.

Nitrogen fixation. Beans and peas nodulate with rhizobia and pull nitrogen from the air, but most of it ends up in the legume itself. What the next crop receives is the residue of roots and stems, a modest contribution that supplements compost rather than replacing it. Beans also compete hard for water while they are fixing.

Trap cropping. A plant the pest prefers, sown earlier than the crop you care about, absorbs the attack and is then removed. That is a different strategy from repelling anything, and it depends on timing.

Two negative mechanisms matter just as much. Competition for water, light, and nutrients is the most common reason a celebrated pairing underperforms. Allelopathy — chemicals released from roots or decaying leaves that suppress neighbors — is the other; fennel and sunflower are the usual garden examples.

A pairing table with reasons

The reason column is where most charts fail, so read it before the claim.

Pairing Usual claim What is happening Verdict
Carrot and onion onion repels carrot fly strong allium scent masks the carrot cue; results vary with density and wind worth a dense row, not a single bulb
Tomato and basil improves flavor no mechanism — flavor is fixed by cultivar, ripeness, and water; basil is a shallow-rooted, harmless neighbor fine, not magic
Corn and pole beans beans feed the corn fixation is largely used by the bean; corn supplies the pole; both compete for water works where corn is a block of 30 or more plants
Lettuce and a tall crop shade and shelter canopy lowers leaf temperature and slows bolting good in July, poor in September
Brassicas and dill repels cabbageworm dill flowers feed parasitic wasps; moths still arrive and lay partial help, cheap to try
Fennel and anything improves growth allelopathic and fiercely competitive; bolts and self-sows keep it out of the vegetable bed

Notice how often the honest verdict is “sometimes.” That is not a failure of the table; it is what a mechanism looks like when it depends on pest pressure, weather, and density. A pairing worth doing is one where the neighbor costs little even if the benefit never shows up.

The marigold question

French marigolds (Tagetes patula) do suppress root-knot nematodes under specific conditions. The effect comes from compounds exuded by roots, it is strongest in sandy soils with a heavy nematode population, it varies by cultivar, and it is largely confined to the root zone during the season the plants are growing. In the trials that show clean results, marigolds are grown as a dense stand for a full season and then turned into the soil — not dotted between tomato plants at 30 cm.

That distinction matters, because the garden-centre version of the advice is a single flat of marigolds along a bed edge, which is unlikely to shift a nematode population by any amount you could measure. As an insect repellent the case is thinner still. There is some support for African marigold (Tagetes erecta) against whitefly in enclosed growing, but nothing that justifies planting them as a general shield.

Marigolds also have costs. They take space a vegetable could use, they can host spider mites in hot dry weather, and the tagetes sold as nematode tools are not the pot marigold (Calendula), which is a different genus with different chemistry. If a soil test confirms root-knot nematodes, grow a solid block of French marigolds for 60 to 90 days and incorporate it. Otherwise plant a few at the bed end because the flowers feed hoverflies, and expect nothing more.

Trap crops and push-pull

A trap crop has to be more attractive than the crop it protects, it has to be in the ground two to three weeks earlier, and it has to be destroyed before the pest completes a generation on it. Skip the third step and you have built a nursery that exports insects across the garden.

Working examples in a home bed: a hillside of blue Hubbard squash sown early pulls squash bugs and cucumber beetles off a later planting of zucchini; a row of radish draws flea beetles away from broccoli, and the radish is pulled and bagged once the beetles gather; nasturtium takes aphids that would otherwise colonize brassicas. Five to ten percent of the bed area in the trap is a reasonable starting point, and the trap needs checking every couple of days, because the whole method depends on intervention.

Push-pull is the more elaborate version, developed for cereal pests in East Africa: a repellent intercrop pushes the insect out of the field while a more attractive border pulls it to a strip where it can be managed. The garden translation is a repellent border around a trap strip, and it works best when the pest is mobile and the pull strip is genuinely preferred over the crop. In a home plot the evidence is mostly practical experience rather than replicated trials, so treat it as an experiment worth one season, not a system to build the garden around.

Spacing and competition

Most charts assume two plants can share a square of soil without cost. They cannot. A tomato’s roots forage 60 to 90 cm out, beans reach 45 to 60 cm, and lettuce spreads 30 cm, so a 30 cm gap between “companions” means fully overlapping root zones and a shared bucket of water and nitrogen.

Height is the other hidden tax. A companion that reaches 1.8 m shades a 30 cm neighbor for much of the morning and evening, when the sun is low. If the low crop needs light, keep 30 to 45 cm between it and anything tall, and put the tall plant on the north side in the northern hemisphere.

The practical rule is adjacency rather than intimacy: put the pairing side by side, not in the same planting hole, and let each plant keep the spacing it would get alone. Rotate plant families on a three-year cycle, because a disease that builds up in the soil does far more damage than a missing companion ever prevents. Flowers work best as 60 cm strips at bed ends or as a dedicated border, at 5 to 10% of the garden’s area.

Pairings that are not worth it

Tomato and potato share late blight and several pests, and both are heavy feeders; the pairing has no upside worth the shared risk. Fennel belongs in the herb border, where its allelopathy and self-sowing do less harm. Garlic or onions next to peas and beans is a common chart entry, yet alliums are widely reported to suppress legume growth, so keep them apart.

The flavor claims deserve a blunt answer. Basil, sage, or marigold beside a tomato does not change the fruit; flavor comes from the cultivar, how ripe the fruit is when picked, and how steadily the plant was watered. Mint near cabbage does not deter caterpillars in any reliable way, and mint spreads by runners fast enough to become the problem instead.

What actually reduces pests is duller and more effective: rotate families, keep 5 to 10% of the area in open flowers, scout the undersides of leaves twice a week, and tolerate a little damage before reaching for anything. Companion planting can sit inside that routine. It cannot replace it.

Sources