"Jack's 321 works for everything." That is how it gets recommended on English growing forums: lettuce, tomatoes, strawberries, cannabis, one answer to every nutrition question. The numbers 3-2-1 are grams per gallon. The recipe is sound, and it is put together carefully. It is also put together for fruiting crops, and on leafy greens it delivers 3.7 times more magnesium than the plant needs. The reason is printed on the label, if you read past the three numbers in the product name.
Quick glossary: Part A (5-12-26) - JR Peters dry blend built from potassium nitrate, magnesium sulfate and monopotassium phosphate, with chelated micronutrients. Part B (15-0-0 Cal Nit) - calcium nitrate, an ammonium calcium nitrate double salt, N 15.0% and Ca 18.0%. "321" - grams per gallon (3.785 L): 3.6 g Part A, 1.1 g Epsom salt, 2.4 g Part B. Epsom salt - magnesium sulfate MgSO₄·7H₂O, carrying 9.86% magnesium and 13% sulfur.
The Recipe, and Why the Order Is Not Optional
The manufacturer gives this as a sequence rather than an ingredient list. Per gallon of water: dissolve 3.6 g of Part A, then 1.1 g of Epsom salt until fully dissolved, and only then add 2.4 g of Part B.
The order is chemistry, not tidiness. The Part B label carries the warning directly: "Mixing Part A and Part B together at high concentrations will result in a messy precipitate - some things just can't go together in the fertilizer world - if this happens, dump and start again." Calcium from Part B and sulfate from Part A give you gypsum; calcium and phosphate give you calcium phosphate. Both are insoluble. The mechanism is the same for any two-part system, and it is laid out in stock solutions A and B.
One useful anchor from the same label: 2.52 g of Part B per gallon, on its own, delivers 100 ppm nitrogen at EC 0.85. The 321 recipe uses 2.4 g, so the calcium nitrate alone accounts for roughly 95 ppm of nitrogen. Part A adds about 48 on top.
The Magnesium Is Already Inside Part A
Part A already carries magnesium and sulfur: the label lists Magnesium 6.3% and Sulfur 8.5%, so 3.6 g of Part A alone delivers 60 mg/L magnesium, over twice the Cornell reference for lettuce before the Epsom salt is even added.
- Magnesium (Mg) 6.3%, water soluble
- Sulfur (S) 8.5%, combined form
- In the "Derived From" list, magnesium sulfate sits second, right after potassium nitrate
Part A is already half Epsom salt by function. At 3.6 g per gallon it delivers 60 mg of magnesium per liter. The Cornell CEA reference solution for lettuce calls for 24 mg/L. Part A on its own overshoots by two and a half times before you open the bag of Epsom salt.
The 321 recipe then asks for another 1.1 g. That is 29 mg/L more. Total: 89.
For contrast: Masterblend 4-18-38 carries 0.50% magnesium, close to nothing, and every bit of it arrives deliberately in a separate bag. Two blends built for one job, assembled back to front from each other, which is why their recipes do not transfer.
What the Recipe Actually Delivers
Against the Cornell CEA lettuce solution, the 321 recipe lands nitrogen and potassium almost exactly on target, but overshoots phosphorus, calcium, and most of all magnesium and sulfur, both roughly 3.5 times the reference.
- Nitrogen: 143 against 125. On target
- Potassium: 205 against 215. On target
- Phosphorus: 50 against 31. Overshoot of 1.6x
- Calcium: 114 against 84. Overshoot of 1.4x
- Magnesium: 89 against 24. Overshoot of 3.7x
- Sulfur: 119 against 35. Overshoot of 3.4x
Nitrogen and potassium land on lettuce almost exactly. The misses sit in magnesium and sulfur, and both arrive from one source: magnesium sulfate, counted twice, once inside Part A and once as a separate bag.
Magnesium at 89 mg/L does not poison anything. The cost shows up elsewhere. First, it is EC spent on an element the crop does not need: the salt is dissolved, the meter reads it, the plant gains nothing. Second, magnesium, calcium and potassium compete for the same uptake pathways, so an excess of one suppresses the others. That is ionic antagonism, and it matters for lettuce because tipburn is a calcium delivery problem rather than a calcium quantity problem. More on that in calcium and magnesium in plant nutrition.
What to Do If You Grow Greens
The simplest correction: drop the Epsom salt entirely on leafy crops. You are left with 3.6 g Part A and 2.4 g Part B, and magnesium falls to 60 mg/L. Still 2.5 times the Cornell figure, but without the double count. To get closer to the reference, reduce Part A and make up the nitrogen with calcium nitrate.
If you grow tomatoes, cucumbers or strawberries, leave 321 alone. Fruiting crops spend magnesium and sulfur differently, and the recipe was assembled for them. Missing on lettuce is misapplication, not a defect in the product.
One more reference point worth holding onto: Cornell runs EC at 1150 to 1250 µS/cm above the source water reading, not as an absolute. Measure absolute EC on hard water and it is easy to underfeed without understanding why. Covered separately in EC in hydroponics.
How this system compares on cost and micronutrient loading against the other two people usually weigh it against is set out in Masterblend vs Jack's 321 vs Flora Series.
Three Mistakes That Cost the Most
Adding Epsom salt "just in case". The standard reaction to interveinal yellowing is to reach for magnesium. If you are running Part A, your reservoir already holds twice the lettuce figure, and the yellowing came from somewhere else: pH, iron, or the root zone. Adding magnesium will suppress calcium and potassium uptake and mask the real cause.
Combining concentrates A and B, or pouring one into the other. Even a splash "to check". Precipitate forms in seconds, the solution still looks clear, and there is now less calcium and phosphorus in it than your calculation says. Always through water, always in sequence.
Treating 321 as doctrine and never measuring EC. The numbers were derived for one water supply and one crop. On hard water, under weak light, or on young transplants, the same recipe produces a different result. A recipe is a starting point; the working value is found with a meter.
How to Tell the Recipe Matches Your Crop
Calculate the magnesium in your own working solution before changing anything: grams of fertilizer times the magnesium percentage from the label, divided by volume. For leafy crops aim at 24 to 40 mg/L; fruiting crops sit higher. If lettuce puts you past 80, you have a double count.
Then read the leaf. Even color, no marginal burn on the young leaves, and steady growth without surges mean the ratios are working. Marginal burn while calcium in the reservoir reads normal points at delivery: humidity, air movement, growth rate, not the dose in the tank.
If you want to go deeper: Nutrient solution recipes: the logic of building one, not copying one covers how to calculate a formula for your own water and crop instead of borrowing someone else's grams.