Why the evidence is thin
Companion planting grew out of gardeners' experience and experiment, not trials. The University of Arizona Cooperative Extension's guide points out why it is hard to study. A garden has too many variables to control, and results often don't repeat from one season or site to the next. An Iowa State University study it describes tested five vegetables with five companion plants. The researchers called their results "indicative", not "conclusive", after unusual weather brought disease and pests. That is typical of the field. The findings below are real, but most come from a single trial, often in the United States or a glasshouse.
Common claims, checked
| Claim | What the research found |
|---|---|
| Marigolds kill nematodes | True only one way. The roots of African and French marigolds produce compounds toxic to root nematodes, but the benefit comes after marigolds are grown as a cover crop and dug into the soil. Interplanting them among the vegetables doesn't give it (Arizona; Mississippi State). |
| Marigolds repel pests in general | Not supported. Mississippi State University Extension says they may not provide the protection gardeners credit them with. |
| Marigolds protect tomatoes from whitefly | Supported in glasshouses. In a 2019 study, French marigolds kept glasshouse whitefly off tomatoes by giving off limonene, and the effect was strongest when they were planted in from the start (Conboy and colleagues, PLoS One). |
| Basil helps tomatoes | Partly. Mississippi State reports studies in which basil and marigolds together reduced thrips on tomatoes. |
| Strong-smelling plants protect cabbages | Some support. Studies reported by Mississippi State found marigolds, nasturtiums or onions planted with cabbage reduced cabbage looper caterpillars. |
| Marigolds and nasturtiums protect zucchini | Some support. In the Iowa State study, marigolds around zucchini reduced squash bug and striped cucumber beetle damage, and nasturtiums reduced squash bugs. |
| Flowers bring in the good insects | Supported. Flowering dill and coriander encouraged predators of Colorado potato beetle and cut damage to eggplants. Sweet alyssum meant more hoverfly larvae and fewer aphids in lettuce (Arizona). |
| Mixing crops confuses pests | Supported in principle. Plants of different heights, leaf shapes and colours can make it harder for some insects to find their target (Mississippi State; Arizona). |
| Legumes feed their neighbours | Partly. Legumes fix nitrogen from the air, and grown as a cover crop or underneath another crop they can reduce fertiliser needs (Mississippi State; Arizona). |
Companions that work for simpler reasons
The most reliable companions work through space, light and timing rather than chemistry. The Arizona guide's examples:
- Fast and slow crops together. Radishes sown with carrots are harvested before the carrots need the room.
- Early and late crops in one bed. Lettuce, spinach or basil grow early, and tomatoes or capsicums go in as they finish. Mississippi State adds that this helps tame weeds.
- Shade where it helps. Taller plants shade smaller ones: tomatoes over carrots, or pole beans shading cabbage to slow bolting.
- Different root depths. Taprooted and tuberous crops such as carrots and potatoes break up compacted soil, and deep-rooted crops draw on water further down.
- The three sisters. Corn is planted first, then pole beans around each stalk once the corn is about 15 cm tall, then squash around the outside. The corn holds up the beans.
Using it in Australian and NZ gardens
Most of these trials used American pests or glasshouse conditions, so treat them as ideas to test rather than guarantees. The approaches most likely to carry over are the general ones. Keep something flowering near the vegetables to feed predatory and parasitic insects such as hoverflies, whose larvae eat aphids. Mix crops instead of planting long single rows. If you want marigolds against nematodes, grow them as a cover crop and dig them in, as the research did, rather than dotting them among the tomatoes.
For "don't plant these together", the pairings with a solid reason behind them come down to disease, not dislike. Crops from the same family share soil-borne diseases, which is what crop rotation is about. The other reason is simple crowding, which the plant spacing calculator helps you avoid.
Questions
Does companion planting actually work?
Some of it. The University of Arizona's extension service calls companion planting "not a proven science". Specific effects do have research behind them: flowers that feed beneficial insects, mixed plantings, some trap and repellent plants, and marigolds as a nematode-suppressing cover crop.
Do marigolds keep pests away from vegetables?
Not in general. Research found specific effects: French marigolds kept whitefly off glasshouse tomatoes, and marigolds reduced some beetles and bugs around zucchini. Against root-knot nematodes they only work grown as a cover crop and then dug in.
Is basil a good companion for tomatoes?
It can help. Mississippi State University Extension reports studies in which basil and marigolds together reduced thrips on tomatoes. Basil also fits in the gaps and is harvested as the tomatoes fill out.
What vegetables should not be planted together?
The best-founded rule is not to keep crops from the same family in the same soil year after year, because they share diseases. Beyond that, avoid crowding and shading sun-loving crops. Most other "bad companion" pairings come from folklore rather than trials.
Sources
- University of Arizona Cooperative Extension, The art of companion planting (April 2023) – the state of the science, the Iowa State study, marigolds and nematodes, dill, coriander and alyssum, space, soil and support companions, the three sisters.
- Mississippi State University Extension, Companion planting: myth or truth? – weeds, legumes, insect confusion, thrips on tomatoes, cabbage looper, the limits of marigolds.
- Conboy N. J. A. and colleagues (2019), Companion planting with French marigolds protects tomato plants from glasshouse whiteflies through the emission of airborne limonene, PLoS One.
Last reviewed 24 September 2026 · Report an error