Reference — updated 2026-07

Sources & Methodology

Every companion planting claim on this site carries one of three stamps. The distinction is the whole point of the project.

A claim is only graded research-backed if we located a controlled trial, extension research, or peer-reviewed study measuring the specific effect. We deliberately do not extend a finding beyond what was tested: the carrot-and-onion intercropping result is graded research for onions and traditional for leeks and chives, because onions are what the study used. Where the literature disagrees, we record the disagreement in the note rather than picking a winner.

1. US Cooperative Extension publications (.edu) — the primary source for growing data and the first place we look for evidence claims.
2. USDA / NOAA / federal agencies — hardiness zones and climate normals.
3. Peer-reviewed literature — for companion planting effects specifically, where extension coverage is thin.
4. Reputable seed companies — used only for spacing and days-to-maturity, never for evidence claims.

Garden blogs, content farms, and unsourced charts are not used as evidence for anything. Where a widely repeated claim traces only to secondary garden sources, it is graded traditional at best, and we say so in the note.

Most sites publish every number with equal confidence. Some of ours are well sourced and some are not, and you deserve to know which is which.

These came up repeatedly during research: extension services publish some things exhaustively and others not at all.

• Mature height. Extension publications almost never state it. Except for a handful of crops (Brussels sprouts and broccoli among them) these are typical-range estimates, not sourced figures.
• Plants per square foot. Essentially no extension body publishes square-foot gardening densities — the sole exception we found names only broccoli. Ours are derived consistently from published in-row spacing (3 in → 16 per square, 4 in → 9, 6 in → 4, 12 in → 1). A sensible conversion, not an independent source.
• Minimum container size and frost-tolerance class. Editorial judgements based on typical root volume and cold behaviour, not published figures.
• Harvest dates in the calendars are days-to-maturity added to the outdoor planting date. Real harvest depends on heat accumulation, so treat them as the earliest plausible date rather than a prediction.

Where a figure could not be sourced and could not be responsibly estimated, we left it out rather than guessing — pepper variety days-to-maturity is one example.

A few things surfaced during research that are worth stating publicly, because they show up uncorrected elsewhere:

• The widely cited "Cornell 65% radish and vine borer study" and a "Nebraska radish study" do not exist as described — the latter is an undergraduate literature review. We left both radish pairings graded traditional rather than myth, since nobody has run the disproving test either.
• Marigold cultivars sold as nematode-suppressing under names that appear in no primary source were dropped from our varieties data. Signet-type marigolds are in fact susceptible hosts that can increase nematode populations.
• Sweet alyssum is a member of the Brassicaceae, so it belongs in the brassica rotation group rather than being treated as a break crop for brassica pests — a detail most companion charts miss.
• One internal contradiction was resolved rather than published: basil originally recorded oregano as a companion on "aromatic reinforcement" grounds, while oregano recorded basil as an antagonist over conflicting water needs. The cultural conflict is far better founded, so both records now agree on it.

As an Amazon Associate we earn from qualifying purchases. Some outbound links to seeds, soil and garden equipment are affiliate links, meaning we may earn a commission if you buy through them at no extra cost to you. We also run display advertising. Neither affects a grade: the evidence stamps are assigned from the literature before any product is linked, and marking a popular product's premise a myth costs us money — which is rather the point of publishing the grades.

Two rules we hold ourselves to on product links. First, we don't recommend anything on a claim we grade as unsupported. The seeds we link on the companion planting pages are the ones behind practices that survived checking — trap crops, insectary flowers — not the pest-repellent products the evidence contradicts. Second, we don't show product photography or prices from the retailer. The illustrations on those cards are our own, and we deliberately don't quote a price, because a cached price goes stale and misleads you within weeks. Click through for the current one.

We do not accept payment for a companion planting grade, a placement in the chart, or a favourable mention, and we will not.

If a figure here is wrong, or you know of a study we missed — particularly one that would move a pairing from traditional to research, or from research to myth — please tell us. Corrections are made to the dataset, which updates every page that uses the figure at once. Records carry the month they were last compiled.

• Monthly: broken links, corrections received, new research on graded pairings.
• Annually, before spring: full review of frost normals and planting offsets.
• On publication: USDA revises the hardiness zone map rarely — the 2023 revision is current as of 2026-07 and superseded the 2012 map.