Companion planting: what the evidence actually shows.

An honest, research-grounded look at which companion-planting claims have scientific support, which are folklore, and the four mechanisms (intercropping, trap cropping, insectary planting, allelopathy) that actually work in home gardens.

Companion planting is one of the most enduring concepts in home gardening and also one of the most loosely-defined. Browse any garden bookstore and you will find dozens of charts claiming that tomatoes "love" basil, carrots "hate" dill, and beans "befriend" everything except onions. Some of these claims have research support. Many do not. Most popular companion-planting charts are folklore copied between books without anyone ever testing them.

This guide separates the two. Per Dr. Linda Chalker-Scott of Washington State University Extension, who has spent two decades reviewing horticultural literature, the popular phrase "companion plant" is too vague to be useful: it is broadly used to describe plant interactions in the realms of science, pseudoscience, and the occult, and the lists that exist describing traditional companion plants have no scientific basis. What does have scientific support is a smaller, more specific set of mechanisms: intercropping, trap cropping, insectary planting, and a few documented allelopathic interactions. This guide covers those mechanisms in detail, explains why some popular claims persist despite lacking evidence, and provides practical guidance based on what the research actually supports.

What companion planting actually means

Per Washington State University Extension guidance, the scientific literature uses several more precise terms instead of "companion planting":

• Intercropping: planting two or more species in the same area, typically in alternating rows, to achieve a benefit that neither species alone provides. Agricultural terminology.
• Polyculture: growing multiple species together in a deliberately diverse arrangement, often with three or more species.
• Trap cropping: planting a sacrificial crop that attracts pests away from the main crop.
• Insectary planting: including flowering plants specifically to attract beneficial insects (predators, parasitoids, pollinators) that benefit the broader garden.
• Allelopathy: the chemical influence one plant exerts on another through compounds released into the soil or air.
• Plant associations: species that are found together in natural environments, often because they share growing conditions or have evolved mutualistic relationships.

When research-literate gardeners talk about companion planting, these are usually the mechanisms they have in mind. When popular gardening books talk about companion planting, the mechanism is often unspecified, the claim is often anecdotal, and the underlying biology is often invented. Recognizing which kind of source you are reading is the first step toward separating useful guidance from folklore.

The four mechanisms with consistent research support

1. Intercropping for niche complementarity

The Three Sisters polyculture (corn, beans, squash) is the most-cited example of intercropping with documented benefits. Per Zhang et al. (2014) in a study published in PLoS ONE, the Three Sisters arrangement produces yield advantages through niche complementarity: corn provides structural support, beans fix atmospheric nitrogen that supports corn growth, and squash vines suppress weeds and reduce soil moisture loss. The three species occupy different above- and below-ground niches and use resources in non-competing ways.

Per Washington State University Extension review of the research, the Three Sisters approach demonstrates "the agricultural practices of polyculture and intercropping, which involve planting mutually beneficial species." The yield advantage compared to corn-only or bean-only or squash-only plots is not as dramatic as popular guides often claim, but it is real and repeatable in controlled studies.

Other intercropping arrangements with research support include fast-growing crops (lettuce, radishes) planted with slower-growing crops (tomatoes, peppers) so the fast crop is harvested before the slow crop fills the space, and deep-rooted crops (carrots, parsnips) paired with shallow-rooted crops (lettuce, onions) that draw nutrients from different soil zones.

2. Trap cropping for pest management

Trap crops are sacrificial plants placed at the edge of a main crop to attract pests away from the primary harvest. Per University of Missouri IPM guidance, blue Hubbard squash planted as a perimeter around summer squash and zucchini effectively traps squash bugs and cucumber beetles, reducing damage to the main crop. The mechanism is well-documented: the pests strongly prefer the trap crop, concentrate there, and either complete their life cycle on a sacrificial plant or are removed manually before they spread.

Other trap-cropping arrangements with research support include nasturtiums for aphids (aphids prefer the soft tissue of nasturtium leaves and concentrate there, where they can be removed), Italian dwarf white sunflowers for stink bugs, and collards for diamondback moths on broccoli and cabbage. The shared feature: a specific pest, a known preferred host, and a deliberate sacrificial planting positioned to intercept the pest before it reaches the main crop.

3. Insectary planting for beneficial insects

Per University of Delaware Cooperative Extension, "scientists have found that there are definite benefits to adding diversity to your garden, primarily because certain plants attract and support beneficial insects that either help control pests or help pollinate your crops." Specific mechanisms include nectar and pollen resources for adult parasitoid wasps (whose larvae control aphids and caterpillars), syrphid flies (whose larvae eat soft-bodied pests), and lacewings.

Per University of New Hampshire Extension research, plants in the carrot family (Apiaceae: dill, fennel, coriander, parsley, Queen Anne's lace) produce small flat-topped flowers that are particularly accessible to small parasitoid wasps and hover flies. Plants in the aster family (Asteraceae: cosmos, calendula, yarrow, sunflowers) and the mint family (Lamiaceae: basil, oregano, thyme, lavender) similarly support beneficial insect populations. A 2024 study published via NIH (Effects of intercropped insectary plants on broad bean predator-pest ratios) documented that sweet alyssum and coriander interplanted with broad beans significantly improved the predator-pest ratio and reduced aphid damage.

4. Allelopathy with documented effects

Allelopathy refers to chemical compounds one plant releases that affect another. Most popular claims about allelopathic companion planting (this plant repels that pest with its scent, that plant suppresses weeds with its roots) lack research support at the practical garden scale. A few interactions have substantial evidence:

• Tagetes patula (French marigold) roots produce alpha-terthienyl, which suppresses certain root-knot nematodes. Per NIH-archived research, this effect is documented and repeatable when marigolds are used as a cover crop (planted thickly for a full season) before susceptible vegetables. The effect is much weaker, and often undetectable, when marigolds are interplanted at low density.
• Black walnut (Juglans nigra) roots produce juglone, which inhibits growth of sensitive plants (tomatoes, eggplant, asparagus) within the dripline. Per Penn State Extension and University of Illinois Extension, this is well-established. Plant susceptible vegetables well away from black walnut trees.
• Brassica cover crops (mustards, radishes used as cover crops) release glucosinolates as they decompose, with documented suppressive effects on certain soilborne pathogens. This is the basis for biofumigation, a research-supported practice in commercial agriculture that has reasonable home-garden application.

Popular claims with no research support

The following claims appear in many popular companion-planting charts. Per Dr. Linda Chalker-Scott's review of horticultural literature and across multiple cooperative extension publications, they lack credible research support:

• "Basil improves the flavor of tomatoes." No mechanism, no controlled studies. Both grow well together because they share similar light, water, and soil preferences. They do not exchange flavor compounds through the soil or air in any measurable way.
• "Carrots love tomatoes." This is the title of a popular gardening book; the underlying claim has no documented mechanism. The two crops coexist fine, but no research has shown that growing them together improves either.
• "Onions repel pests from carrots." A small handful of studies has tested this; the consistent finding is that onions do not significantly reduce carrot fly damage when interplanted.
• "Tomatoes hate brussels sprouts" / general 'hate lists'. Per WSU Extension's review, "There is no scientific basis for any of the several lists that exist describing 'traditional companion plants.' Like horoscopes, these lists may be fun to use, but they should not be perceived or promoted as scientifically valid."
• "Plant chamomile near anything to make it grow better." Chamomile is a fine garden plant. Claims that it confers general benefit to neighboring crops are unsupported.

Per the University of Illinois Extension companion-planting review, "Much of the recommended companions that we see are not always tested out in a research study. They may be more anecdotal." This is the honest framing: anecdote is not evidence, but anecdote can be a useful starting point for hypothesis-testing. If your particular garden has produced reliable results from a specific pairing, that experience has value; just do not assume it generalizes.

What this means in practice

The Three Sisters, properly

The Three Sisters (corn, beans, squash) deserves specific mention because it is the most-cited example of companion planting and one of the few with substantial research backing. Per the Zhang et al. (2014) study and corroborating research, the arrangement works through three documented mechanisms:

• Corn provides physical structure for climbing beans, eliminating the need for trellising.
• Beans fix atmospheric nitrogen via root nodule symbiosis with Rhizobium bacteria, providing some nitrogen to the system (research has shown the amount is modest in a single season but real over multi-year cycles).
• Squash vines provide ground cover that suppresses weeds, retains soil moisture, and discourages some pests through physical interference.

The traditional Indigenous planting method (per multiple ethnobotanical sources) involves planting corn first, allowing it to establish for two to three weeks, then planting beans at the base of established corn plants, and finally planting squash between the corn-bean clusters. The squash spreads to fill the open ground between plant clusters as the corn and beans grow. The arrangement requires a fair amount of space (a 10-by-10-foot bed is roughly the minimum to see the dynamic work) and adequate sunlight.

For most home gardeners, the practical version is closer to a polyculture bed: plant a small block of corn, surround it with pole beans, and let squash or pumpkin vines spread underneath. The yield advantage compared to growing the three crops separately is modest but real, and the structural and weed-suppression benefits are immediate.

The takeaway

Companion planting in the strict sense (specific pairings that confer specific benefits) has uneven evidence support. Some claims have research backing. Many do not. The popular charts are mostly folklore, and the gardeners who follow them with mixed results are typically experiencing the noise of variable garden conditions rather than the signal of specific plant interactions.

What does have research backing is the broader principle that diverse gardens outperform monocultures. The mechanisms are well-understood: more flower diversity supports more beneficial insects, more structural diversity creates more microhabitats, more root depth diversity uses soil more efficiently. None of this requires a specific charm about which plant "loves" which. It requires planting variety, including insectary flowers, considering trap crops for documented pest problems, and avoiding the mistake of treating folk lists as scientific protocols.

Diverse is better than monoculture. Specific magic pairings are mostly folklore. The mechanism is the message: intercropping, trap cropping, insectary planting, and the few documented allelopathic interactions are the actual tools that work.

Plant variety. Include flowers. Use trap crops for specific problems. Treat the popular charts as a starting point for experiments, not a protocol. Your garden will do better, and you will understand why.