Short answer. A solid mineral feed for cannabis in coco or hydro runs roughly N 150–200, P 30–60, K 175–250, Ca 150–200, Mg 40–70, S 60–120 mg/L, plus Fe 1.5–3 mg/L and traces of Mn, Zn, B, Cu and Mo. Run it at about 1.2–1.8 mS/cm in veg and 1.6–2.2 mS/cm in flower. Cannabis trials show no benefit from pushing P above about 60 mg/L or from doubling overall strength [4][5][6]. Below: the targets, the evidence, and two worked recipes for RO water.
Targets by stage
All values mg/L (= ppm by mass of the element) in the solution going to the roots, including what your water supplies. EC values are total solution EC at 25 °C.
| Element | Seedling / clone | Veg | Transition (flower wk 1–3) | Mid flower (wk 4–7) | Late flower / ripen | Evidence |
|---|---|---|---|---|---|---|
| EC (mS/cm) | 0.6–1.0 | 1.2–1.8 | 1.6–2.2 | 1.6–2.2 | 1.2–1.8 | Grower practice; 4 vs 2 mS/cm gave no gain [6] |
| N | 50–100 | 150–200 | 160–200 | 140–190 | 100–150, tapering towards 0 at the finish (see note) | Veg optimum ~160 [1]; flower optimum ~194 [2]; taper is Grower practice |
| NH₄ share of N | ≤10% | ≤10–20% | ≤10–20% | ≤10–20% | ≤10% | >30% harmful [8] |
| P | 15–30 | 30–50 | 40–60 | 40–60 | 30–50 | No gain above ~25–60 [4][5][6]; optimum ~59 in DWC [2] |
| K | 60–120 | 175–240 | 180–250 | 200–280 | 180–250 | Veg 175–240 [3]; no flower response 60–340 [2] |
| Ca | 80–120 | 150–200 | 150–200 | 150–200 | 130–180 | Standard solutions [9][10]; Supported |
| Mg | 25–40 | 35–70 | 40–70 | 40–70 | 40–60 | Veg 35–70 [7] |
| S | 30–60 | 60–100 | 60–120 | 60–120 | 60–100 | Supported (standard solutions) |
| Fe | 1.0–2.0 | 1.5–3.0 | 1.5–3.0 | 1.5–3.0 | 1.5–2.5 | Greenhouse standards [9] |
| Mn | 0.2–0.5 | 0.3–0.8 | 0.3–0.8 | 0.3–0.8 | 0.3–0.6 | [9] |
| Zn | 0.1–0.3 | 0.1–0.5 | 0.1–0.5 | 0.1–0.5 | 0.1–0.4 | [9] |
| B | 0.2–0.4 | 0.2–0.5 | 0.2–0.5 | 0.2–0.5 | 0.2–0.5 | [9]; toxicity at ~5 mg/L [10] |
| Cu | 0.03–0.05 | 0.03–0.1 | 0.03–0.1 | 0.03–0.1 | 0.03–0.08 | [9] |
| Mo | 0.01–0.05 | 0.01–0.05 | 0.01–0.05 | 0.01–0.05 | 0.01–0.05 | [9] |
| Na, Cl | As low as possible; Na <50, Cl <50–70 | See sodium |
Notes: - Per-stage splits within veg and flower are grower practice; the research gives single optima per phase, mostly from one or two cultivars per study. - Organic liquid feeds need higher N on paper (one coir trial put the flowering optimum at 212–261 mg/L N) because organic N is less available [see organic vs mineral]. - Soil growers: these are solution numbers; soil supplies and buffers its own nutrients. - Feed through to harvest. The late flower / ripen column is a real feed, not a wind-down. Don't flush before harvest: pre-harvest flushing is a Myth that starves the plant while it packs on its final weight, trichomes and resin, so it costs you yield and resin [see the flush debate]. The only valid flush is a root-zone flush to clear salt build-up, at any stage, when runoff EC says you need one (see runoff testing). - N and potency are a trade-off. Lower N in flower can raise cannabinoid and terpene percentage while lowering total yield; yield rose with N up to ~160 mg/L in one controlled trial (Saloner & Bernstein 2021, Industrial Crops and Products 167:113516, https://doi.org/10.1016/j.indcrop.2021.113516). Pick your late-flower N deliberately within the table range; it isn't a reason to flush (see nitrogen). - Taper N, don't flush. From about 3 weeks out, ramp N down: ~75 % of late-flower N at week −3, ~50 % at −2, ~25 % at −1, 0–20 mg/L in the final days, with P, K, Ca, Mg, S and micros unchanged (Grower practice). That needs single salts with an N-free Ca source (see hydroponic salts) or per-element dosing; diluting a 2-part cuts everything.
Sources table: which salt to use for each element
| Element | Main sources | Secondary sources |
|---|---|---|
| N (nitrate) | Calcium nitrate, potassium nitrate | Magnesium nitrate, nitric acid |
| N (ammonium) | Already in calcium nitrate (~1.1% NH₄-N) | MAP, ammonium nitrate (small amounts) |
| P | MKP | MAP, phosphoric acid |
| K | Potassium nitrate, potassium sulfate | MKP, potassium silicate, KOH |
| Ca | Calcium nitrate | Calcium chloride (small amounts) |
| Mg | Magnesium sulfate | Magnesium nitrate |
| S | Magnesium sulfate, potassium sulfate | Sulfuric acid |
| Fe | Fe-DTPA (hydro, coco), Fe-EDDHA (alkaline) | Fe-EDTA (pH < 6.5) |
| Mn, Zn, Cu | Sulfate salts or EDTA chelates | |
| B | Boric acid | Sodium borate |
| Mo | Sodium molybdate | Ammonium molybdate |
Salt percentages are in hydroponic salts explained.
Worked recipe 1: veg, RO water
Per 1,000 L (divide by 10 for 100 L). Assumes RO water with EC < 0.05 mS/cm.
| Salt | g per 1,000 L | Supplies (mg/L) |
|---|---|---|
| Calcium nitrate (15.5% N, 19% Ca) | 840 | N 130, Ca 160 (NH₄-N ≈ 9) |
| Potassium nitrate | 380 | N 53, K 147 |
| MKP | 175 | P 40, K 50 |
| Magnesium sulfate heptahydrate | 500 | Mg 49, S 65 |
| Potassium sulfate (44.9% K) | 50 | K 22, S 9 |
| Totals | N 183, P 40, K 220, Ca 160, Mg 49, S 74 |
Worked recipe 2: flower, RO water
| Salt | g per 1,000 L | Supplies (mg/L) |
|---|---|---|
| Calcium nitrate | 790 | N 123, Ca 150 (NH₄-N ≈ 9) |
| Potassium nitrate | 200 | N 28, K 77 |
| MKP | 220 | P 50, K 63 |
| Magnesium sulfate heptahydrate | 560 | Mg 55, S 73 |
| Potassium sulfate | 240 | K 108, S 44 |
| Totals | N 150, P 50, K 248, Ca 150, Mg 55, S 117 |
Micronutrients for both
| Salt | g per 1,000 L | Supplies (mg/L) |
|---|---|---|
| Fe-DTPA (11% Fe) | 23 | Fe 2.5 |
| Manganese sulfate monohydrate | 1.6 | Mn 0.5 |
| Zinc sulfate heptahydrate | 1.3 | Zn 0.3 |
| Boric acid | 2.3 | B 0.4 |
| Copper sulfate pentahydrate | 0.3 | Cu 0.08 |
| Sodium molybdate dihydrate | 0.1 | Mo 0.04 |
Weigh micros into a 1:1000 stock (for example, 1,000× the per-litre amount in 1 L of water) so you dose millilitres, not milligrams.
Expected EC. A rule of thumb for nutrient solutions is EC (mS/cm) ≈ sum of cations in meq/L ÷ 10 [9]. Recipe 1 has about 18 meq/L of cations (Ca 8.0, K 5.6, Mg 4.1, NH₄ 0.7), so expect roughly 1.8 mS/cm; recipe 2 ≈ 1.9 mS/cm. Measure; don't trust the arithmetic alone.being reviewed
How to use it
- Get a water analysis and subtract what the water supplies (Ca, Mg, S, Na, Cl, alkalinity). See water quality.
- Mix A and B stocks separately (calcium nitrate and iron in A; everything else in B).
- Fill the tank, add A, stir, add B, stir, then adjust pH to 5.6–6.0 (hydro) or 5.8–6.2 (coco).
- Check EC against the target.
- Monitor runoff or reservoir (see runoff testing) and adjust strength, not ratios, first.
- Tissue test once per cycle if you can (most recent mature leaf): N 3.3–5.0%, P 0.27–0.48%, K 1.8–2.7%, Ca 1.5–2.9%, Mg 0.30–0.65%, S 0.25–0.36% (NC State floral hemp ranges).
Common mistakes
- Copying a feed chart's EC without accounting for tap water EC.
- Switching to a high-P, low-N "bloom" feed at the flip and starving the stretch.
- Stacking boosters on a complete base (see bloom boosters).
- Assuming more EC means more yield: doubling to 4 mS/cm gave no yield or quality benefit in deep-flow hydro [6].
Evidence
- N, P, K, Mg optima: Supported (controlled cannabis trials, mostly one or two cultivars each [1][2][3][4][5][7]).
- P and overall-strength ceilings: Established as "no benefit above" across several independent trials [4][5][6].
- Ca, S and micronutrient targets: Supported by general greenhouse standards [9] and solutions used in cannabis research [10]; no cannabis dose-response data for most.
- Per-stage EC ladder: Grower practice.
