Which Microgreens Should a Beginner Grow? The Two Crops on Every List Are the Two This Directory Does Not Rate Beginner

Five beginner pages were opened and audited. Radish, pea shoots and sunflower appear on all five. Broccoli appears on four. That is close to unanimous, and it is the part of this subject with the least evidence behind it.

No extension publication found says "these are the crops for a first-time grower" and backs a ranking with comparative outcome data. Utah State tells a home grower to choose microgreen-labeled seed "such as" radish, broccoli, sunflower or pea shoots, which is an example list rather than a tested recommendation. Maryland calls the cabbage and mustard family especially well suited. Neither establishes which crop makes the best first tray.

So this directory's own fields are the closest thing to a check that exists, and they disagree with the consensus on exactly two crops.

The consensus set, against this directory's own ratings

Pea and sunflower are the only two of the five that this directory does not rate as beginner crops, and they are the two that appear on every single list.

They are also, on every mechanical axis, the two with more to go wrong:

• Both need a soak, 6 to 8 hours for pea and 4 to 12 for sunflower, while radish, broccoli and mustard need none. A soak is a step a first-time grower can forget, mistime or run too warm.
• Both carry high mold risk in this directory, against medium for the other three.
• Both sow heavy. Pea at 150 g and sunflower up to 167 g a tray, against broccoli's 13 g and mustard's 8 g. This directory's median seed rate is 23 g, so the two universal picks sow at roughly 6.5 to 7.3 times the median crop, and pea uses 18.8 times mustard's seed.

That last one is the money. A failed first tray of mustard costs 8 grams of seed. A failed first tray of pea costs 150. The sowing page owns the full 409-fold seed-rate range, and this is the corner of it that lands on a beginner.

None of this says pea or sunflower fail more often, because nobody has measured that. It says the two crops recommended most universally are the two carrying the most steps, the most seed and the higher mold rating, which is the opposite of what "forgiving" is supposed to mean.

Seventeen beginner-rated crops are never named at all

This directory rates 62 crops for difficulty: 20 Beginner, 33 Moderate, 9 Difficult.

Three of those 20 are on the consensus list. The other 17 are named by none of the five pages: arugula, broccoli raab, cauliflower, clover, collard, garden cress, green cabbage, kale, kohlrabi, komatsuna, lentil, mizuna, pak choi, red cabbage, tatsoi, turnip and wheatgrass.

And 18 of the 20 beginner-rated crops need no soak at all, including most of that unnamed list. If the fewest steps is the criterion the beginner pages say they are using, the pages are not applying it.

One number that surprised me and is worth publishing. Zero of the 20 beginner-rated crops carry a low mold risk. Not one. That is not a gap in the data, it is the answer: mold is a property of your tray, not of your crop. Maryland and Penn State both describe it that way, naming moist rather than saturated media, airflow and vented covers as the controls, and warm humid conditions as what favors damping off. No extension source found assigns a comparative mold ranking by crop, and the practice pages that do offer one supply no tray counts, conditions or comparator.

What "easy" actually means, and what nobody has measured

There is no published operational definition of an easy crop. These are the criteria the sources actually state, and what each is worth:

• Fast germination. Maryland puts cabbage-family germination at about 36 hours at 72°F and says the family germinates faster than the beet family. That is speed, not forgiveness, and they are not the same thing.
• Fewer special steps. Maryland lists pea, sunflower, corn, chickpea, beet, chard, cilantro, mung bean and buckwheat for a 6 to 12 hour soak. A no-soak crop removes a task. No study tests whether removing it makes a novice more likely to succeed.
• Tolerance of low light. Maryland says lettuce, arugula, mustards and Asian greens grow acceptably under less-than-ideal light. This is the closest thing to an extension-backed forgiveness criterion that exists, and note that it names almost none of the consensus crops.
• Something you want to eat. Clemson and Utah State both publish flavor tables: broccoli mild, sunflower nutty, pea sweet, radish nutty and spicy, arugula peppery. A technically perfect tray nobody eats is a failed recommendation.

What has never been measured is the thing the word actually means. No study found deliberately imposed the same novice mistakes, overwatering, missed watering, weak light, temperature swings and uneven sowing, across common crops and compared how many trays stayed usable. No source defines easy as the probability of a usable first tray, and none reports that probability.

Practice pages use outcome language without a denominator. "Usable harvest within two weeks" is not a success rate if nobody counts the failures.

Two studies do report per-crop germination percentages, and this page quotes neither. Both returned a 403 to a direct fetch, so their individual values could not be verified at source. A number you cannot open is a lead, not a citation. In any case germination is batch, temperature and treatment specific: Penn State's seed-density tool asks for the germinability of your own batch, which is the correct framing and the reason a permanent per-crop ease score does not exist.

Why five pages agreeing is not five pieces of evidence

All four practice lists share the same core. Every one contains radish, pea and sunflower; three of four add broccoli; two add mustard. Exact agreement is zero of four, because list lengths and extras differ.

All four give reasons, and none gives evidence. The reasons offered are speed, vigor, no soak, large harvest, flavor and claimed tolerance. Not one links its specific "forgiving" assertion to a comparative novice trial, because none exists.

That is a shared recommendation tradition, not independent derivation. It is not proof of copying either. It is the ordinary situation where everyone draws on the same familiar core and nobody publishes failure data, and the honest reading is that agreement here carries much less weight than it looks like it should.

The pages get real things right, and dismissing them would be the opposite error. They separate flavor and purpose from growing mechanics, which matters. They warn that pea and sunflower carry an extra soak. They warn against mixing species in one tray, which Maryland and Penn State both support because germination and growth rates differ. Practice observations generate useful testable warnings. They just must not be restated as measured mold rates or success probabilities.

One more thing they get right that this directory confirms: a retailer's product link is not a sales ranking, and none of these pages publishes units or revenue by crop. What crops beginners actually buy and grow first has never been surveyed.

What to actually do

What nobody has measured

• A novice usable-harvest trial. New growers randomly assigned one crop and a simple protocol, with "usable" defined in advance as safe, non-moldy, harvestable and actually eaten. This is the single most valuable missing study in this subject.
• Crop against imposed mistake. A replicated trial running normal, over-watered, under-watered, low-light, temperature-swung and unevenly sown treatments across crops, reporting usable harvest.
• A comparative mold ranking by crop under identical seed density, water, temperature and airflow, with diagnosed disease and tray-level incidence.
• Whether large seeds are genuinely easier to sow, measured as sowing time, density error, skipped steps and usable harvest, rather than assumed from visibility.
• Whether a short crop fails less often. Fewer days is less time for something to go wrong, and that has never been tested against actual losses.
• What beginners actually grow first, from a prospective survey recording crop, setting, prior experience, harvest and whether they ate it.
• Whether seed cost changes persistence after a first failure, which is the mechanism behind the 8-gram against 150-gram point above and is entirely untested.

Sources

Opened 2026-08-12. The difficulty, mold risk, soak, seed rate and harvest figures for individual crops are this site's own variety fields, read across all 62 entries; the counts of 20 beginner, 33 moderate and 9 difficult, the 17 unnamed beginner crops, the 18 beginner crops needing no soak and the zero beginner crops at low mold risk are computed from those fields rather than taken from any source below. The list counts are from the five pages named here, audited on one date, and are not a count of every beginner page online. Two studies reporting per-crop germination percentages were opened and are deliberately not quoted, because a direct fetch returned 403 and their individual values could not be verified at source.

• Utah State University Extension, grow your own microgreens - the 20 February 2025 guidance naming radish, broccoli, sunflower and pea shoots as example microgreen seed choices, the dated crop table giving pea a 6 hour and sunflower a 12 hour soak with radish and broccoli at none, covered time from 1 to 2 days for radish to 7 to 14 for cilantro, a 2 to 5 lb weight over the first half of the dark period, flavors including arugula peppery, broccoli mild, pea sweet, radish nutty and spicy and sunflower nutty and sweet, and the note that repurposed pie tins or takeout trays can work.
• University of Maryland Extension, growing microgreens and baby greens indoors - the cabbage and mustard family named as especially well suited and germinating faster than the beet family, about 36 hours at 72°F for cabbage-family germination, the 6 to 12 hour soak list covering pea, sunflower, corn, chickpea, beet, chard, cilantro, mung bean and buckwheat, lettuce, arugula, mustards and Asian greens named as growing acceptably under less-than-ideal light, seed counts of 8,000 per oz for broccoli and 15,000 for mustard against 1,500 for beet, the advice against mixing species because germination and growth rates differ, and moist rather than saturated media with airflow and vented covers as the disease controls.
• Clemson Extension, growing microgreens - the 21 November 2018 guidance on a shallow 1 to 2 inch tray with drainage, keeping light about 4 inches above new sprouts to guard against leggy seedlings, avoiding garden soil, manure and compost and using untreated microgreen-labeled seed, the definition of microgreens as harvested when true leaves begin to form, and the harvest-time and flavor table covering broccoli, sunflower and pea among others.
• Penn State Extension, growing microgreens - the advice to select one type per tray, and damping off identified as the only disease encountered in its class with warm humid conditions favoring it.
• Penn State Extension, the ABCs of microgreens - the seed-density tool asking for the germinability of the grower's own batch, which is why a permanent per-crop germination score does not exist.
• Practice, five beginner pages audited 2026-08-12 as a record of what is recommended rather than as evidence, and the source of the list counts: Urban Harvest Lab, listing radish, pea, sunflower, broccoli and mustard with a reason for each and warning that mucilaginous chia, cress and basil are hard to spread and mold-prone; Wind River Greens, listing sunflower, pea and radish, warning that crowded wet sunflower can mold and that cilantro needs cracking or pre-hulling, and using outcome language without a denominator; Garden UK, listing pea, radish, sunflower and broccoli and using "taste good enough" as an explicit starter criterion; and CityRooted, an affiliate page ranking on speed, flavor, yield and cost. None publishes crop-level sales figures, and none links a forgiveness claim to a comparative trial.
• Opened but deliberately not quoted, because a direct fetch returned 403 and their values could not be verified at source: Plants, 2024, assessment of vegetable species for microgreen production in unheated greenhouses, Agriculture, 2023, evaluation of growth conditions, antioxidant potential and sensory attributes of six diverse microgreens species, and Agronomy, 2025, annual garden rocket and radish as microgreens, seed germination response to thermal and salt stress.