Companion Planting 101: What Actually Works (and What's Folklore)

Type almost any two vegetables into a search bar with the word "companion" between them and a chart appears within seconds, stating with total confidence whether the pairing helps or hurts. Follow enough of those charts upstream and they thin out fast - most trace back to one of a small number of gardening paperbacks from decades ago, republished as web copy so many times over that the original reasoning, if there ever was any, is impossible to recover from the version you're reading now. This page draws the line the charts won't: not folklore versus nonsense, but folklore versus evidence, which turns out to be a very different and far more useful distinction.

The Actual Problem With Companion-Planting Charts

The charts fail in a specific, nameable way, and it's the reason this whole site exists. They print every claim in the same confident tone - one row for "corn supports climbing beans," one row for "marigold repels pests," one row for "basil improves tomato flavor" - as though the three sat on equal footing. They don't. One of those three is basic physics with a plant attached. One is real, but far narrower than the row implies. One has never been demonstrated to happen at all, by anyone, in any controlled setting we're aware of. A reader scanning the chart has no way to tell which is which, because the chart's format doesn't allow for the distinction. That's the actual failure: not that folklore gets repeated, but that folklore and evidence get printed in identical type, at identical confidence, next to each other.

Corn, Beans, and Squash: The One System Where Every Leg Checks Out

Most tested companion claims involve one crop affecting one other crop through a single mechanism. The planting method refined over centuries by Indigenous farmers of North America is built differently - it stacks three separate, independently explainable relationships into one mound, which is a large part of why it survives scrutiny better than almost anything else covered here. A mature corn stalk gives pole beans something to climb, functioning as a living trellis that costs nothing and needs no maintenance beyond letting the corn get a head start. Squash's broad, low leaves spread across the ground between the mounds, shading out weeds and slowing moisture loss the way any dense ground cover would, occupying a different vertical layer than the corn so neither crop shades the other out.

The nitrogen part is where people usually round off a corner. Bean roots host bacteria that fix atmospheric nitrogen into a usable form, and a portion of that does become available to a nearby heavy feeder like corn while both are still growing - real, but modest, since most of the nitrogen a bean plant fixes stays locked in its own root and leaf tissue until that tissue breaks down. The larger nitrogen payoff usually lands on whatever gets planted in that same soil the following season, not on corn sharing the bed with the bean plant right now, which is exactly why the benefit belongs in rotation planning as much as in this season's mound.

The corn-and-bean trellis relationship has its own page covering that mechanics in more depth, and the squash-and-corn ground-cover pairing gets its own look at the shading effect. For the mound layout and spacing itself, see our Three Sisters planting guide, and for the fuller case on why this particular three-crop system holds up better than most companion claims, we've written that argument out separately in The Three Sisters: Why Corn, Beans and Squash Actually Work Together. The follow-on nitrogen question connects directly to how rotation works in a small garden - the beans' real payoff often lands there, not in this year's mound.

Onions and Carrots: A Two-Way Effect Somebody Actually Measured

If this site had to defend one pairing to a skeptic, it would be onions and carrots, because the trial record here is unusually direct. Carrot rust fly hunts by scent, and onion's sulfur-heavy aroma genuinely interferes with its ability to find carrot foliage. The reverse holds too: onion fly and onion thrips locate onions by smell, and carrot foliage's own scent disrupts that search in turn. What separates this from most of the claims on this page is that field trials compared intercropped plantings directly against each crop grown alone and measured a real reduction in pest damage on both sides - not a related mechanism borrowed from a different context, but the actual pairing, actually tested.

It isn't total protection - heavy regional pest pressure can still get through - but it's rare enough to find a pairing with two-way trial data behind it that we'd put carrots and onions near the top of anything on this site actually worth building into a bed plan, not just filing away as an interesting claim.

Giving Pests Something Better to Eat, on Purpose

Trap cropping gets confused with scent-masking constantly, and the two are opposites. Scent-masking tries to make a crop harder to find. A trap crop does the reverse - it advertises a more attractive target right next door and accepts the damage that follows as the cost of protecting what actually matters. Nasturtium interplanted with brassicas is the clearest documented case: aphids show a consistent, field-observed preference for nasturtium foliage over kale, cabbage, or cauliflower leaves, pulling colonies onto a plant you were never counting on for a clean harvest anyway. A trap planting that looks pristine all season isn't necessarily succeeding - it may just mean the aphids went straight for your brassicas instead, which is worth knowing before you judge whether the tactic worked.

We've dedicated a full page to untangling trap cropping from masking, since mixing the two up tends to produce a planting choice that doesn't actually satisfy either strategy - see Trap Cropping Explained. The specific brassica-and-nasturtium relationship, including how much space to give the trap planting, is covered on the kale-and-nasturtium pairing page.

Marigold: Real, But Not for the Reason Most People Plant It

No plant gets recommended for pest control in a home vegetable garden more often than marigold, and it earns roughly half of that reputation honestly. French marigold's roots release thiophene compounds - alpha-terthienyl among them - that are genuinely toxic to root-knot nematodes, a real, repeatable, soil-science-confirmed effect. The catch most charts leave out is scale: the studies behind that effect grew marigold as a dense stand across an entire plot for a full season, long enough for the nematicidal compound to accumulate in the soil before anything else went into that ground. A handful of marigolds tucked around an already-established tomato plant this season isn't reproducing that setup, even if it isn't doing nothing either.

The claim most gardeners actually believe when they plant marigold - that its scent keeps whitefly and aphids off nearby crops above ground - runs on a completely separate mechanism, and it's the half of marigold's reputation that doesn't hold up under the same scrutiny. Scent-masking against flying pests is plausible in principle and confirmed for some other pairings, but no controlled comparison we know of has singled out marigold and shown it actually cutting above-ground pest damage on a neighboring crop. Marigold's real strength works underground, on one narrow category of pest, through root chemistry - not as an all-purpose aromatic bodyguard for whatever happens to be planted nearby.

The full breakdown of the nematode research versus the whitefly folklore lives on the marigold-and-tomatoes pairing page, and general growing details for the plant itself are on its crop guide and planting-timing page.

The Tomato-and-Basil Question Nobody Has Actually Tested

Few pairings show up in as many vegetable gardens as tomatoes and basil, and the two justifications gardeners give for growing them side by side don't rest on anything like equal footing. The idea that basil's scent throws off egg-laying moths hunting for a tomato to target isn't far-fetched on its own - other documented pairings show an aromatic plant interfering with exactly this kind of host-finding by smell - but nobody appears to have isolated and confirmed the effect for basil against the specific moths responsible for tomato hornworm eggs. The idea that basil changes how a nearby tomato fruit tastes is a separate, far more popular claim, and it has essentially nothing behind it: there's no known pathway for a neighboring plant's aromatic oils to travel through the soil and into a tomato's flavor chemistry. This is the culinary world's favorite garden pairing riding on a claim its own kitchen use never actually needed.

None of that makes the pairing pointless - basil works as a genuinely useful low, shading ground cover at a staked tomato's base, and the two want nearly identical sun and water. The full case for planting them together anyway, and why we still don't rate the flavor claim, is on the basil-and-tomatoes pairing page.

Fennel Is the One Case Where "Keep Them Apart" Has Real Chemistry Behind It

Fennel is the rare case where a "these two don't get along" warning survives contact with actual chemistry. The plant is allelopathic - its root and leaf tissue carries anethole and related essential-oil constituents that leach into the surrounding soil and suppress germination and growth in nearby species, tomatoes included, reliably enough across separate studies that gardening sources have stopped treating the effect as speculative. Because the suppression travels through the soil itself rather than depending on the plants touching, there's no useful middle ground of "just a bit closer than usual" the way there is with most antagonist warnings on this site - fennel effectively needs its own footprint, in a dedicated bed or a large container, kept well away from nightshades in particular.

The specific research and how far to actually separate the two plants is covered on the fennel-and-tomatoes pairing page.

When Two Nightshades Share a Disease, Not Just a Last Name

Tomatoes and potatoes aren't kept apart because they're distantly related and gardening lore is suspicious of family reunions. Both are hosts for the pathogens behind early and late blight, and both attract the Colorado potato beetle, so growing them close together simply shortens the distance a real outbreak has to travel from one crop to the other. That's plant pathology, not a vague family-based caution, and it's why what happened in a bed last year matters more here than this year's spacing chart - once blight has shown up in a plot, both crops need to stay well clear of that ground for several seasons, not just clear of each other this one.

One soil pathogen ignores that family logic completely. Verticillium wilt infects an unusually wide host range spanning both nightshades and cucurbits with no regard for how distantly the two are related botanically, which is why a cucumber planting can pick up a caution flag near potatoes despite having no family connection at all. The fungus also persists in bare soil without needing a living plant to survive on between seasons, so one clean rotation year rarely clears a bed that's had an outbreak - treat it as a multi-year problem rather than a one-season fix.

Both mechanisms, and the practical rotation timing that follows from each, are laid out on the potatoes-and-tomatoes and cucumbers-and-potatoes pages.

Alliums Turn on Legumes, Even Though They Help Almost Everything Else

Garlic and onion are helpful neighbors almost everywhere else on this site, which makes this exception worth flagging on its own. Their sulfur-based root exudates interfere with the rhizobia bacteria a legume's roots need to form nitrogen-fixing nodules, and research looking specifically at allium-legume interaction has measured reduced nodulation and slower nitrogen fixation in beans and peas grown too close to onions or garlic. This is a scheduling problem more than a spacing one: give a bed at least a full season before following an allium with a legume in the same ground, rather than trying to manage it through distance within a single season.

The research and the root-architecture reasoning behind it are covered on the bush-beans-and-garlic pairing page.

The Antagonisms That Don't Get a Catchy Name

Two more real conflicts rarely make it onto a chart because they're less dramatic than a named disease or a chemical compound, but they're genuine and worth knowing. Some brassicas release glucosinolate breakdown products into the soil as their residue decomposes, and while this can suppress certain soilborne pests, the same residue left to break down directly around a following crop's root zone can measurably slow germination in sensitive neighbors - a reason to compost spent brassica material rather than till it straight back into the same bed where something delicate is about to be sown. And the plainest antagonism of all needs no chemistry at all: two heavy feeders sharing tight quarters simply run out of the same nutrients at the same time, showing up as stunted growth that has nothing to do with pest confusion or root exudates and everything to do with a bed that can't support two hungry crops' demands simultaneously. It's the least interesting explanation on this page, and it's also quietly behind more disappointing companion-pairing results than any chemical mechanism on this page ever gets credit for.

One more entry worth naming for what it isn't: two crops from the same plant family, like carrots and parsnips, aren't a real threat to each other the way fennel or Verticillium are - they just share a pest and a root zone, which is a missed opportunity rather than a hazard. See the carrots-and-parsnips page for the distinction, and our roundup of What Not to Plant Together: The Real Antagonists for the shortlist of conflicts we're confident enough in to call genuinely worth planning around.

Why We Don't Just Delete the Unproven Half of This Site

It would be easy to read all of the above and conclude that anything short of "evidence-backed" should be removed. We don't think that follows, and the reasoning matters more than it looks. There's a mundane cause behind why so much of this has never been tested in a controlled setting: replicated intercropping studies are slow and expensive to run, and there's essentially no funding lined up to answer a question like whether basil changes a nearby tomato's flavor. A missing study tells you the question hasn't been answered yet, not that the answer is no - and a home gardener risking nothing but a few seeds and one growing season is in a very different position than a research budget deciding what's worth financing. Treating "nobody has checked" as the same thing as "proven wrong" would mean claiming a certainty this site doesn't actually have.

This is why every pairing on this site carries an explicit evidence-backed, mixed, or traditional-lore label rather than one uniform tone. We've written out the full reasoning behind each label, and what separates them, on Evidence vs. Lore: How We Rate Every Companion-Planting Claim - read that page for the method; this one is where we apply it across the whole field at once.

Putting This Into an Actual Bed

None of this needs to turn planning a garden into a research project. The practical takeaway is simpler than the mechanisms behind it: lean on the evidence-backed pairings when you're choosing what genuinely matters for pest pressure or yield, treat traditional lore as low-risk experiments rather than settled fact, and take the handful of real antagonists seriously enough to actually separate or rotate around them, because those are the ones with a documented downside if you ignore them.

• Start from a crop's own growing needs, not just its companion list - our individual crop guides and planting-timing pages cover sun, spacing, and frost tolerance that matter more to a plant's success than any companion pairing does on its own.
• Check a specific pair before committing bed space to it rather than trusting memory or a printed chart - our companion pairing checker tool cross-references any two crops against the full rated dataset in seconds.
• If your garden is small, don't force in a sprawling system like the three sisters just because it's the most famous one - a compact layout built around your actual bed size will outperform a folklore favorite that doesn't fit the space.
• Time any pairing against your own last frost date, not a generic calendar month - two crops can be a great match on paper and still fail together if one goes in weeks before local conditions are ready for it.

For that last point specifically, see Reading Your First Frost Date. For layout, Square-Foot Gardening is the compact alternative worth reading if a full three-sisters mound doesn't fit your space. And for checking any pair directly, use the Companion Pairing Checker.

What This Page Can't Tell You

This is a planning guide, not a diagnosis. It can tell you what's worth interplanting before you plant, and what to keep apart, but it can't look at a real leaf with real spots on it and tell you what's actually gone wrong once a season is already underway.

If something you've already planted is struggling right now - curling, discoloring, or just failing to thrive - that's a different question than anything on this page, which is written for the planning stage before a seed goes in the ground. Our sister site WhyMyPlant is built for that after-the-fact moment, walking through what a struggling plant is actually telling you rather than what to interplant it with.

For everything else - the full catalogue of individual pairings, every crop's own growing guide, and the planting-timing layer that ties both to your actual calendar - the companion pairing hub, the crop guide hub, and the when-to-plant hub are the three starting points everything on this page ultimately points back to.