Pak choi microgreens

This directory runs on two kinds of number, and until now they have never been checkable against each other on the same crop.

One kind is a commercial trial: Johnny's weighed 29 varieties in 1020 flats and published grams in and ounces out. The other is a controlled study: researchers grow a crop at a stated density in square meters, and this site converts the result by area. Four entries here rest on that conversion, and nobody has ever been able to say whether it is any good.

Pak choi is the crop where both exist. Johnny's measured 10 and 11 oz per 1020 at first true leaf. A controlled study, independently, measured 2.10 kg per square meter at the same stage, which area-converts to 9.56 oz per 1020. That is 4 to 13 percent apart, from unrelated methods, on the same crop at the same harvest stage.

That is the most reassuring result this directory has produced, and it is about the method rather than the vegetable.

What it is, and how to tell you grew it right

Pak choi microgreens are seedlings of Brassica rapa subsp. chinensis. Pak choi, bok choy, pac choi and Chinese cabbage are market names for the same thing, not four crops.

**That makes it the fourth Brassica rapa in this directory, alongside komatsuna, mizuna and tatsoi, still to be written. The komatsuna entry says plainly that these siblings cannot be separated at cotyledon stage and this page does not contradict it. What pak choi does have is a stem: the hypocotyl is notably thicker and juicier than its siblings', and the cotyledons are broad, smooth and spoon-shaped where mizuna's first true leaf turns angular and toothed. That is a tendency and not a key. No controlled photographic comparison of green B. rapa types at cotyledon stage exists**, so keep the seed label.

A correct tray is an even carpet at two to three inches early or seven to ten centimeters at the later target, with thick pale-white to pale-green stems and paired smooth cotyledons. Red selections carry pigment in stem and leaf; a generic green bag will not.

What it tastes like, and what to do with it

Mild, faintly sweet Chinese cabbage with a gentle brassica edge and no heat at all. The flavor is carried by texture as much as taste: the stem is crisp and full of water, and that is the point.

So it goes where a dish wants crisp juicy volume rather than a seasoning. Pile it on finished ramen or pho, fold it into a noodle bowl off the heat, use it as a salad base, or scatter it as a fresh garnish. Brief residual heat softens the stem without destroying it.

It is the wrong choice when the plate needs a peppery hit, which radish or mustard supplies, or a feathery herb texture, which mizuna does better.

People who want a spicy microgreen will find it too mild, and people who dislike cabbage flavors will still notice the brassica note. No consumer study exists, so the honest test is a sample at your own harvest stage.

Growing it: only what differs from the standard method

Everything general is on the standard tray method for growing microgreens. Pak choi's deltas are few, which is most of why it is a beginner crop.

Sow dry at 10.5 g, no soak. Two crop-specific guides say no soaking is needed, and the seed is small and round rather than dust-fine, so it meters evenly by hand. The trial rate is 10.5 g for both measured cultivars.

Three to four days covered, with a light weight and only during germination. The published figure is 1.1 to 2.2 lb evenly distributed, for root anchoring and straight vertical stems rather than for compression. Off when most seedlings are pushing the tray above them.

Then get the air moving before the canopy closes. This is the one instruction that matters on this crop. A thick, water-rich stem in a dense stand traps humidity at the medium line, and the crop-specific guide flags root-neck rot with stagnant air specifically. Bottom water from the moment roots are established.

Seed lower, not higher, if you are growing past first true leaf. Johnny's own guidance is that a larger plant needs fewer seeds, for airflow and slower depletion of water and fertility. The temptation on a crop this productive is to crowd it for a heavier cut, and that is the failure mode rather than the optimization.

Hold 60 to 70°F. The crop guide reports usable performance at 61 to 72°F. Johnny's observed that high-80s and low-90s conditions accelerated growth while damaging uniformity and finished quality, which is a more useful warning than a simple ceiling: the tray will look fast and finish badly.

What goes wrong with pak choi specifically

Uneven emergence in the first three days, from unevenly distributed seed or unevenly wet medium. Clumps and dry gaps. Not repairable; accept the grade or resow.

Stem-base rot, and this is the characteristic failure. Excess moisture, poor air exchange or a crowded stand. The early tell is specific and worth memorising: the stem at the medium line looks translucent, water-soaked or pinched before anything falls over. Caught there, better drainage and airflow can save it. Once seedlings collapse at the medium line it is not food. Pythium and Rhizoctonia are the usual organisms.

On root hair against mold, pak choi makes this harder than most crops because a healthy tray produces conspicuous white root fuzz. Root hair is bright white, symmetrical, concentrated at the roots and the medium line, and it visibly shrinks when misted. Mold is patchy or web-like across seed, stems and medium, may run gray or green, smells off, and does not vanish when wet. Location plus the mist test, not whiteness.

Thin pale stretched stems from weak light or too long under cover. Correctable going forward; the height and uniformity you already lost do not reset.

Cut late and it goes woody. Stems coarsen, leaves yellow, true leaves advance and the tray loses uniformity. The controlled study found that waiting from first to second true leaf raised fresh yield from 2.10 to 2.80 kg/m², a third more weight, and its authors are explicit that this does not establish a better product: harvest stage needs a genotype-specific cost-benefit judgement. That is the same trade komatsuna records at 26 percent, from the same experiment.

Nutrition, and what is actually measured

There is real crop-specific measurement here, and it is the mineral panel.

The controlled study grew a red pak choi genotype at 70,000 seeds per square meter, germinated dark at 24°C, and harvested at both true-leaf stages with three replicates. Among the four Brassica rapa genotypes it compared, pak choi came out highest in phosphorus, potassium, magnesium, iron and zinc: phosphorus 10.58 and potassium 17.08 mg per gram dry weight, iron 180 and zinc 141 µg per gram dry weight.

Two things about that are worth stating plainly. It is a dry-weight panel from one experiment, not a per-serving figure, and converting it to what is on a plate requires a moisture figure the study is being used carefully about here. And it is a comparison among four siblings, which is exactly the useful form: it does not tell you pak choi is nutritious in the abstract, it tells you that among the B. rapa types this one carries the most of five minerals.

What is not established is a USDA panel. No FoodData Central record for pak choi microgreens was confirmed, and mature bok choy values are not a substitute.

So the supplier claims quoting precise vitamin A, C, K, calcium, folate and glucosinolate figures have no analytical basis attached, and "many times more than mature pak choi" is unverified for every nutrient, cultivar and harvest stage. The measured panel above is better than any of that, and it is narrower than any of that.

Which seed to buy

Three things are sold and two of them have numbers behind them. Rosie F1 is a named red hybrid with a measured 10.5 g and 11 oz. Red Pac Choi is a separate trial entry at 10.5 g and 10 oz. Generic "pak choi microgreen seed" names no cultivar, which is fine for a green mild crop and supports no promise about yield, stem length or color.

Specify untreated seed for edible microgreens, with a lot number, test date and germination percentage. Refuse coated or fungicide-treated seed. Johnny's own trial excluded flats with obvious poor germination from its averages and warns that seed lot changes yield, so the published 10 to 11 oz is a benchmark rather than a guarantee.

The marketplace is the better route for one tray and the worse route for ten. The Amazon search returns 177 results with eight genuine pak choi microgreen listings: Kitazawa Dwarf White Stem 1 oz at $15.39, Shanghai 1 oz at $17.73, and small bags at $3.98 for 10 g and $4.56 for 2 g. That 10 g bag is roughly one tray for four dollars. Johnny's works out at $2.50 a tray but only if you buy a pound at $107.85.

One listing on that page is worth flagging: "Pak Choi Rosette Tatsoi". Pak choi and tatsoi in a single product title, on a shelf where the two cannot be told apart in a tray. That is the fourth sibling-labeling trap this directory has recorded, after the komatsuna group, the onion alliums and the nasturtium name.

Is it worth growing?

Yes, and it is one of the strongest recommendations in this directory, which is not something many crops here earn.

It returns 11 ounces in 13 days from 10.5 grams, which is 0.85 ounces per tray-day, fourth in the whole directory behind sunflower, radish and broccoli and just ahead of mizuna. At $2.50 of seed that is 4.4 ounces per dollar, which is mid-pack rather than good, and the reason is the seed price rather than the rate.

What sets it apart from the other mild brassicas is the stem. Kale, cabbage, komatsuna and mizuna are all chasing the same dense-tender-mild slot; pak choi is the one that arrives with crunch and water in it, which is why it holds up in a noodle bowl where the others wilt. Add a red cultivar and it is also the one that looks like something.

Choose something else if you want heat, a second cut, or the cheapest possible tray. And if you are buying seed by the ounce rather than the pound, know that the tray is closer to $5 than $2.50.

A note on the numbers above

The corroboration is the headline and it is worth being precise about. The study measured 2.10 kg per square meter at first true leaf. A 1020 is 0.129 square meters, so that is 271 grams, or 9.56 oz. Johnny's, in an unrelated greenhouse with its own medium and its own seeding, measured 10 oz and 11 oz at the same stage. The conversion lands 4 percent under one and 13 percent under the other.

That matters beyond this crop. Four entries here publish an area-converted figure because it is the only measurement that exists: komatsuna, fennel, lentil and onion. Every one of those cells is labeled as a conversion and every one of them has carried an unstated question: does the arithmetic survive contact with a real tray? On the one crop where the comparison can be made, it does, to within about a tenth. That is not proof the method transfers everywhere, and edge effects on a study tray a fifth the size of a 1020 are a real reason it might not. It is one datum, and it is the only one anybody has.

Pak choi and komatsuna come out of the same experiment, which is the other thing to say plainly. This directory has been caught before treating two figures from one document as independent corroboration when they were one source repeating itself. Here the study's pak choi figure is corroborated by Johnny's, not by komatsuna, and the two pages should be read as sharing a source rather than confirming each other. Both now say so.

On bench turnover pak choi is fourth in the directory at 0.85 oz per tray-day, behind sunflower's 1.79, radish's 1.21 and broccoli's 0.92, and just above mizuna's 0.83. On value it is mid-pack at 4.4 oz per dollar, and that is the seed price rather than the crop: at $107.85 a pound Rosie is expensive for a brassica, against mizuna's $18.76.

The pack-size ladder on this crop is unusually flat and that is worth knowing. Most crops here move threefold between a retail bag and a bulk one; mung bean moves from $5.48 to $1.80 and nasturtium from $10.07 to $2.75. Pak choi at 25 lb is $2.22 against $2.50 at a pound, an 11 percent saving. Buying in bulk barely helps here, which is unusual and mildly good news for a small grower.

The harvest window is blank, which as of 2026-08-09 covers 45 of the 49 entries written.

All 61 crops, sortable, on the microgreen comparison.

Four of the crops above carry an area-converted figure and pak choi is the reason you can trust them a little more than you could yesterday. Komatsuna, fennel, lentil and onion all publish a number that came out of a square meter. On this crop that arithmetic can be checked against a real tray, and it holds.

Sources

• Johnny's microgreens yield trial results (PDF) - the measured rows: Rosie at 10.5 g, 11 oz and 13 days, and Red Pac Choi at 10.5 g, 10 oz and 13 days, both to first true leaf.
• Johnny's yield trial, method and discussion - 1020 flats, at least three seedings per variety, flats with obvious poor germination excluded, the seed-lot warning, and the guidance to seed lower for a larger plant.
• Johnny's Rosie F1 bok choy - Brassica rapa var. chinensis, and the price ladder verified 2026-08-09 at packet $6.45, 1 oz $13.25, 1/4 lb $33.90, 1 lb $107.85, 5 lb $519.80 and 25 lb $2,395.75.
• Kyriacou et al., 2021, Brassica rapa microgreens at two harvest stages (PDF) - the controlled study: 70,000 seeds/m², dark germination at 24°C, n=3, pak choi at 2.10 kg/m² at first true leaf and 2.80 at second, highest P, K, Mg, Fe and Zn of the four genotypes, and the authors' point that harvest stage needs a genotype-specific cost-benefit assessment. The same study behind the komatsuna entry's figures.
• UA Organic, pak choi microgreens - the crop-specific guide: no soak, 3 to 4 days dark with a 0.5 to 1 kg weight, 16 to 22°C, 16 to 18 hours of light, bottom watering, the root-neck rot warning under stagnant air, the abundant white root hairs, cutting by day 11, and the water-rich stems making a pack run wet.
• BageechaBox, pak choi microgreens - the 8 to 10 day claim at 18 to 28°C and 2 to 3 inches, small round seed, no soak, and the mild sweet flavor description.
• Virginia Tech Extension SPES-756 (PDF) - bok choy as B. rapa var. chinensis, uncovering after majority germination at 2 to 5 days, the 60 to 70°F and 50 to 70% RH baseline, and dark storage at 41°F.
• SDSU Extension, growing microgreens at home - the general method, the roughly one week refrigerated figure, and the advice to buy untreated seed.
• UC IPM, damping-off - Pythium and Rhizoctonia, and seed and seedling death at or below the medium line.