Myth vs. Reality: Companion Planting

Most companion planting charts and guides are written as though every pairing carries equal authority — basil-with-tomato sits next to marigold-with-everything sits next to corn-beans-squash, all presented with the same confident shrug. The conventional wisdom is that this is fine, because companion planting is "folk knowledge" and folk knowledge does not need controlled studies. The conventional wisdom is wrong. Some pairings have measurable, peer-reviewed yield and pest-suppression data behind them — Three Sisters polyculture documents a 20-40% yield advantage over monoculture, sweet alyssum insectary strips deliver 50-70% aphid reductions, and French marigolds, planted with very specific protocol, reduce root-knot nematodes by 40-60%. Other pairings, sold with equal confidence in every chart on the internet, fall apart the moment anyone runs a controlled trial. This guide grades the difference, plant by plant.

Quick answer: does companion planting actually work?

Some of it does, some of it doesn't, and the easy way to tell the difference is to ask whether the claim has a measured effect size attached. Trap cropping, insectary planting, intercropping, and certain cover-crop sequences are peer-reviewed and quantified. Most of the pairings in Louise Riotte's 1975 Carrots Love Tomatoes — the source of the popular companion-planting canon — were never tested in a controlled trial and don't survive the ones that have been run since. This chart grades every pairing with ✅ Peer-reviewed, ⚠️ Anecdotal-but-promising, or ❌ Folklore so you can sort the difference yourself.

The 7 science-backed companion planting strategies

The University of California Master Gardeners of San Mateo & San Francisco have done the cleanest work consolidating what actually has evidence behind it. "Better Together: The New Science of Companion Planting" organises the evidence-backed practices into seven distinct strategies — none of them folklore, all of them with measurement. The list is more useful than any chart, because it tells you why a pairing works rather than just asserting that it does.

1. Trap cropping — plant a more attractive sacrificial crop near the one you want to protect. Blue Hubbard squash draws cucumber beetles away from your zucchini and cucumbers; nasturtium draws aphids off brassicas and tomatoes. University of Missouri IPM and University of Nebraska-Lincoln trials document 30-60% reduction in pest pressure on the protected crop.
2. Masking — strong-scented herbs interfere with pest cues. This is the mechanism behind "basil near tomato" — not flavor improvement, which has no biological basis, but volatile-organic-compound interference that disrupts whitefly and aphid host-finding. Effect size is small but measurable.
3. Insectary plants / living mulch — a 5-10% allocation of bed area to flowering insectary species (sweet alyssum, dill, fennel, yarrow, buckwheat) delivers 50-70% aphid reductions and 2-4x more beneficial insects, per UNH 2023 trials and USDA ARS. This is the single most actionable evidence-backed intervention on the list.
4. Intercropping — different species, different vertical layers, different ecological roles, sharing one bed. Three Sisters (corn, beans, squash) is the textbook example, with a Land Equivalent Ratio of 1.2-1.4 — meaning 20-40% more total yield per unit area than the same species grown separately.
5. Soil conditioning — deep-rooted plants like daikon radish bring up nutrients from below the rooting zone of shallower crops, and break compaction. The benefit is delayed-release — it accrues to the next season's planting, not the concurrent one.
6. Disease suppression — French marigolds (Tagetes patula) suppress root-knot, lesion, and reniform nematodes via the root compound alpha-terthienyl. The protocol is restrictive (see the marigold section below); the effect, under the right conditions, is real and documented.
7. Pollinator support — flowering companion plants that overlap the bloom window of the crop you want pollinated. Useful for squash, cucumbers, and any crop that depends on insect pollination for fruit set.

Everything else in popular companion-planting media — "basil improves tomato flavor," "carrots love tomatoes," "mint repels everything," the entire mid-century catalog — is either ⚠️ anecdotal-but-untested or ❌ debunked. Three of the seven strategies above (insectary plants, trap cropping, marigold-as-cover-crop) deliver enough effect size to justify the bed space on their own. The rest of this guide is the chart you can use to make decisions and the deep-dives on the two pairings that get systematically misreported in every other gardening article on the internet.

Loading image...

The evidence-graded companion planting chart

Use this chart as the working reference. The Evidence column is the part no other chart on the internet has, and it is the part that should change what you actually plant.

Pairings marked ⚠️ are widely reported and not actively contradicted — useful but not proven. Pairings marked ❌ in the Why column (legume nitrogen transfer in the same season; "basil improves tomato flavor"; mint as a general pest deterrent) appear in many other guides as fact and should be treated as folklore until controlled trials say otherwise.

Companion plants for tomatoes (in detail)

Tomatoes are the most-searched companion-planting query on the internet — and the most-misrepresented. The two claims you will see in every chart are "basil improves tomato flavor" and "marigolds repel tomato pests." Both are partly wrong.

Basil-near-tomato does have an evidence base, just not the one most charts cite. The flavor claim is biologically implausible — Garden Myths cataloguer Robert Pavlis is direct that "plants have limited ability to absorb complex flavor molecules through the air or soil". What basil actually does is mask host-finding cues for whiteflies and certain aphids by releasing volatile organic compounds that disrupt their search. The effect is small but measurable. Plant basil at the base of tomatoes because it shares the same water and light preferences and gives a modest pest-cue interference — not because it makes tomatoes taste better.

Lettuce planted in the shade of mature tomato plants delays bolting in summer heat. This is a mechanically real benefit, well-supported in extension publications, and the easiest companion planting to demonstrate in a backyard garden — same lettuce variety, two beds, one in tomato shade and one in open sun, the tomato-shaded bed will hold for two to three weeks longer in late June.

Marigolds and tomatoes are a more complicated story; the protocol below covers it in detail. The short version: marigolds as a cover crop the season before tomatoes go in is supported. Marigolds interplanted with tomatoes is not.

Avoid fennel near tomatoes — fennel is one of the few vegetables with documented allelopathy, releasing compounds that inhibit a wide range of garden neighbors. Avoid corn, because corn earworm is the same species as the tomato fruitworm, and they will move between the two crops happily. Avoid brassicas, because they're heavy nitrogen feeders that will compete with tomatoes and shade emerging plants in early summer.

Companion plants for cucumbers

Cucumbers benefit from the single best-documented trap-crop pairing on the chart: nasturtium for aphids and cucumber beetles. University of Missouri IPM trials and University of Nebraska-Lincoln work document 30-60% reductions in pest pressure on the protected crop when nasturtium is planted along the edges of a cucumber bed. The mechanism is straightforward — the pests prefer nasturtium and concentrate there, where you can either tolerate the damage on a non-food crop or remove the infested nasturtium leaves wholesale.

Dill flowering at the edge of the bed pulls in aphid predators (lady beetles, lacewings, parasitic wasps) and is part of the insectary-plant strategy. Radishes interplanted in the row act as a flea-beetle trap. Corn is the Three Sisters partner — see below. Avoid sage and rosemary, which prefer drier soil than cucumbers tolerate, and avoid brassicas, which compete for the same shallow-bed nutrients.

Companion plants for peppers and onions

Peppers share most of their tomato pairings — basil at the base for cue masking, spinach in the cool-season window, onions at the edges. Avoid beans (both peppers and beans suffer in proximity for reasons not fully characterised but consistently reported) and brassicas (the heavy-feeder competition story).

Onions are the cheapest win on the chart, ranking at KD 2 in keyword difficulty and supported by genuine effect-size data. Onion volatiles disrupt carrot rust fly host-finding, which is the cleanest documented allium-companion effect. Carrots planted in alternating rows with onions show measurably less fly damage. Pair onions with carrots, beets, brassicas, and tomatoes. Avoid beans and peas — the allium family is allelopathic to legumes, and the effect is reproducible.

Lettuce and broccoli

Lettuce gets shade from tall companions (tomatoes, peppers, pole beans) and delays bolting in summer; this is the most easily reproduced companion-planting demonstration on the list. Add radishes and chives as quick-growing fillers between lettuce rows. Avoid brassicas, which compete heavily.

Broccoli and the brassica family benefit from dill at the edge — a textbook insectary species that pulls in the parasitic wasps and lacewings that depress aphid populations. Onions and beets work as physical neighbors; celery is the traditional pairing. Avoid tomatoes and strawberries near brassicas due to shared-pest overlap.

The Three Sisters: corn, beans, and squash

Three Sisters is the canonical example of evidence-backed companion planting and the one every gardener should grow at least once, even on a small scale. The mechanism is genuinely beautiful — three species, three vertical layers, three ecological roles, one bed.

Corn provides the vertical structure. Climbing beans use the corn stalks as a living trellis, eliminating the need to build one. Beans, via Rhizobium bacterial symbiosis in their root nodules, fix atmospheric nitrogen into a plant-available form — but the catch most articles miss is that the benefit accrues to next season's planting, not the concurrent corn and squash. Isotopic tracer studies show only 2-8% nitrogen transfer between legumes and neighboring plants during co-cropping; the real nitrogen return happens after the bean residue decomposes. Squash leaves spread across the soil surface, forming a living mulch that suppresses weeds, retains soil moisture, and creates a microclimate friendlier to beneficial insects.

The yield advantage is documented as a Land Equivalent Ratio of 1.2-1.4 — meaning 20-40% more total yield per unit area than the same three crops grown separately as monocultures, citing Zhang et al. 2014 and Mead & Willey 1980. That is a genuinely large effect size for a no-input intervention.

To plant Three Sisters in a backyard, you need at least a 4-foot-square bed (corn requires depth for wind support). Plant corn first, in a 4×4 grid pattern. Wait until corn is six to eight inches tall — this is non-negotiable, because beans planted simultaneously will out-grow the corn and topple it. Then plant pole beans at the base of each corn stalk. Plant squash or pumpkin around the edges last. Water consistently; the squash mulch will reduce frequency once it fills in.

Marigolds: the protocol nobody else publishes

This is the section that justifies the entire "myth vs reality" frame. Every chart on the internet says to "plant marigolds near tomatoes to repel nematodes." The chart is almost right — and the difference between "almost right" and "useful" is the difference between forty dollars of marigold seedlings doing something and forty dollars of marigold seedlings doing nothing.

The science is clear enough that the University of Florida IFAS Extension publication NG045 is unambiguous about it. French marigolds (Tagetes patula) suppress root-knot nematodes (Meloidogyne species), lesion nematodes, and reniform nematodes via a root compound called alpha-terthienyl — described in the literature as "one of the most toxic naturally occurring compounds found to date" against these pests. The 40-60% nematode reduction over an 8-12 week protocol is genuine, USDA-NRCS-cited, and reproducible.

The catch is the protocol. UF IFAS states the conditions plainly: marigolds must be planted at the exact location where the susceptible vegetable will go, at less than 7-inch spacing, as a dense cover crop for at least two months before the vegetable is planted. The susceptible vegetable then goes in after the marigolds are pulled or worked into the soil. And UF IFAS is explicit that "intercropping marigold with other crops to reduce plant-parasitic nematodes does not appear to be effective" — citing Powers et al. 1993, which tested marigold interplanted with cucurbits and found less productive crops than the monoculture, with no nematode suppression. Every other gardening article on the internet tells you to interplant marigolds with tomatoes. The most rigorous Extension publication on the topic says that doesn't work.

Variety matters too. The resistant Tagetes patula cultivars — 'Toreador', 'Alaska', 'Crackerjack', 'Single Gold', 'Tangerine' — are the ones the published trials used. The Signet marigolds ('Golden Gem' and 'Tangerine Gem'), which are widely sold as garden ornamentals, are themselves susceptible to root-knot nematode and can amplify rather than suppress the problem. A gardener who plants Signet marigolds expecting nematode control is buying the wrong plant for the right reason.

The honest summary is this: marigolds are a useful tool against root-knot nematodes if you treat them as a pre-vegetable cover crop in the right spot at the right density for the right duration with the right cultivar. As a generic "plant a few near your tomatoes" companion, the evidence does not support the claim. Above-ground pest deterrence — the broader "marigolds repel garden pests" claim — has been tested and is largely unsupported in controlled trials.

Loading image...

What NOT to plant: antagonistic pairings

The "what to avoid" column on the chart deserves its own short section because it captures real allelopathy and shared-pest mechanisms.

• Tomato + fennel. Fennel releases allelopathic compounds that inhibit a wide range of garden neighbors. The pairing is among the few "avoid" recommendations with documented mechanism.
• Tomato + brassica. Heavy-feeder competition for nitrogen plus partial shade incompatibility.
• Tomato + potato. Both are Solanum species and share blight susceptibility — late blight (Phytophthora infestans) moves between them readily.
• Tomato + corn. Corn earworm and tomato fruitworm are the same species (Helicoverpa zea) and will move between adjacent crops.
• Bean + onion (and other alliums). Allium-family allelopathy on legumes is documented and reproducible. Plant separately.
• Strawberry + brassica. Shared aphid and beetle pest overlap.
• Carrot + dill (flowering). Dill, once flowering, cross-attracts pests; carrots are best planted away from mature dill.

Sample bed layout

If you have one 4×8 raised bed, here is a layout that works on actual evidence rather than tradition. The bed is divided into thirds.

• First third (4×3): tomatoes (3-4 plants), with basil at the base and a row of carrots along one edge. Lettuce filling the partial-shade corner.
• Middle third (4×3): brassicas (broccoli or cabbage, 2-3 plants), with dill at the edge for the insectary effect and onions along the row. Sweet alyssum as a 6-inch border strip.
• Final third (4×2): Three Sisters or a cucumber-and-nasturtium pairing, depending on whether you have height. Sweet alyssum at the edge again.

The sweet alyssum border around two of the three sections is the single most evidence-backed companion-planting choice you can make in this layout — 5-10% of the bed area for 50-70% aphid reduction is a return that almost nothing else in the garden delivers.

Loading image...

The one thing to plant this weekend

If you take one thing from this guide: plant a strip of sweet alyssum at the edge of every vegetable bed you own. Total build: a $3 packet of seed, ten minutes of seeding, no holes in anything. 5-10% of the bed area, 50-70% aphid reduction, two to four times the beneficial-insect population. That single intervention out-performs every folkloric pairing in every chart on the internet. Three Sisters in May, marigolds as a pre-tomato cover crop in spring of next year, nasturtium between the cucumbers in June. Skip the rest until the controlled trials catch up.