Most of What You Have Read About Companion Planting Came From One Book Published in 1975
I spent two growing seasons religiously following a companion planting chart I had printed from a gardening website. Tomatoes with basil. Carrots away from dill. Marigolds around everything. Beans are never near onions. The chart had the confidence of established science, the kind of information that gets presented as settled fact with no ambiguity, and I followed it accordingly, adjusting my planting decisions around it every spring.
It was only when I started reading more carefully about where these recommendations actually came from that I found something surprising. A significant portion of modern companion planting folklore, particularly the specific plant pairings that appear on many charts in circulation, was popularised by a book called Carrots Love Tomatoes, published in 1975 by Louise Riotte, though some traditions predate it. The book was charming, enthusiastically written, and largely based on traditional garden lore and personal observation rather than controlled experiments. It became enormously influential. Most of the companion planting charts you find today are downstream of it, and most of the specific claims in those charts have never been formally tested to see whether they hold up.
What Testing Actually Reveals
When researchers have studied specific companion planting claims under controlled conditions, the results are more complicated than the charts suggest. The basil and tomatoes pairing, probably the most famous in companion planting, has been tested in controlled studies and has not produced consistent evidence of a growth or flavour benefit to the tomatoes from basil's presence. Both plants grow well under similar conditions, which may be where the association came from: if you grow them together and both do well, it's natural to assume the proximity is helping. Under conditions where the variable is genuinely controlled and only the presence or absence of basil changes, the tomatoes perform about the same either way.
The marigold-and-nematode claim is one of the most frequently cited examples of companion planting that works, and the reality is more specific than the simple version suggests. French marigolds do produce chemical compounds that suppress certain soil nematodes. Still, the effect requires growing marigolds as a full-season crop in the affected area and then incorporating them into the soil, not tucking a few marigold plants around your tomatoes. The scattered-marigolds-as-pest-repellent approach that most companion planting charts recommend has not been shown to produce the nematode suppression that the research demonstrates is possible under different conditions.
Beans fixing nitrogen for neighbouring plants is another claim that sounds logical and is true in the wrong sense. Legumes do fix atmospheric nitrogen through bacteria in their root nodules, but that nitrogen becomes available to neighbouring plants only after the legume roots die and decompose. Growing beans next to this season's tomatoes does not feed the tomatoes this season. Leaving bean roots in the soil at the end of the season, or growing a legume cover crop and incorporating it before planting, does enrich the soil for subsequent crops. The mechanism is real. The timeline in the popular version of the claim is not.
What Actually Does Work
Trap cropping is the most consistently evidence-backed companion planting strategy and the least glamorously presented. The principle is simple: plant something that a pest prefers more than your crop, positioned to draw the pest toward the decoy plant rather than the one you want to protect. Nasturtiums genuinely attract aphids strongly enough that a border of them around brassicas can concentrate aphid populations on the nasturtiums rather than on the cabbage, at which point you remove the nasturtium plants along with the aphids before they spread. It requires active management rather than passive benefit, which is probably why it appears less frequently on companion planting charts than the more appealing idea of plants passively protecting each other.
Planting flowers among vegetables to attract predatory and parasitic insects is also genuinely useful, with the important clarification that the benefit comes from the diversity and presence of flowering plants generally rather than from specific magical pairings. Sweet alyssum attracts hoverflies and parasitic wasps that also consume aphids. Borage and calendula attract pollinators. The specific plant pairings matter less than having a range of flowering plants present throughout the season. A diverse planting generally produces a healthier garden than monoculture rows, independent of any specific companion planting pairing.
The Three Sisters combination of corn, beans, and squash is one of the few traditional companion planting systems where the benefits have been observed consistently and where the mechanisms are well understood. The corn provides structure for the beans to climb, the beans fix nitrogen that benefits the corn, and the squash leaves shade the ground to suppress weeds and retain moisture. It's a genuinely functional system developed over thousands of years of agricultural practice, and it is meaningfully different from most companion planting charts in that each element provides a documented, structural benefit to the others rather than a vague chemistry-based improvement that has never been measured.
Why I Still Plant Marigolds
None of this means a companion planting chart is useless or that the practice has no place in a home garden. It means that most of the specific claims deserve more scepticism than they typically receive, and that the decisions worth making are the ones where the mechanism is understood rather than where the pairing has simply been repeated often enough to seem authoritative.
I still plant marigolds around my beds, not because I have evidence they are keeping specific pests away from my tomatoes, but because they are easy to grow, attract beneficial insects generally, and look good in a predominantly green space. That is a perfectly sufficient reason to grow them. It's just a different reason from the one on the chart, and being clear about the difference means I am not making planting decisions based on a claim that has not actually been tested. Good soil preparation and the right growing conditions consistently do more for plant health than any specific pairing, which is not as interesting a claim as the idea that certain plants are friends and enemies, but it has considerably more evidence behind it.
Conclusion
The companion planting chart I followed for two seasons was not based on decades of agricultural research. It was based largely on one enthusiastically written book from 1975, which passed through several decades of repetition until it acquired the authority of established fact. Some of what is in those charts is genuinely useful, some of it works under specific conditions that the chart does not describe, and some of it has not been tested at all. Treating the specific pairings as reliable rules to follow is giving them more credit than they have earned. Treating the general principle, that a diverse garden with flowering plants present throughout the season tends to be healthier than a monoculture, as a useful guide, is giving it about the right amount.
