The Estimative Index is a method, not a recipe: add more of every nutrient than the plants could possibly use — roughly 23 ppm of nitrate, 3.9 of phosphate and 23 of potassium across a week, in 3 doses — then throw 50 % of the water away every week so none of it can accumulate. You never test, because the estimate is deliberately generous and the water change catches the error. And the point of all that is not the plants: it is that fertilizer stops being one of the things that could be wrong.
Almost every explanation of EI you will find is a schedule. Dose this much on Monday, that much on Tuesday, change half the water on Sunday. The schedule is easy and it is on a dozen pages, so this one starts somewhere else: with what the method is for, why it was invented, and the two questions the forum threads actually argue about — whether the excess is necessary, and whether you can skip the water change. For the grams themselves, that is a number, and numbers live in the fertilizer dosing calculator.
What the Estimative Index actually is
It is a dosing method published by Tom Barr in the early 2000s, and it has three parts that only work together:
- Dose every nutrient into excess. Not "enough" — more than the plants can take up, so that no nutrient is ever the thing holding growth back.
- Dose on a rhythm, not on a reading. Macros a few times a week, traces on the days between, the same amounts every week regardless of what the tank looks like.
- Reset with a large water change. A big weekly change hauls the accumulated excess back out, which is what makes the first part safe to do.
Barr's original write-up gave weekly ranges rather than doses — nitrate 20–30 ppm, phosphate 1–3, potassium 20–30, iron 0.5–1 — on the reasoning that plants cannot use more than that and a large water change caps whatever they do not. The familiar "7.5 ppm three times a week" set that every chart prints is a way of hitting them, not the method itself, and it does not hit them exactly: it lands inside the range for nitrate, potassium and iron and above it for phosphate, at 3.9 ppm a week against a published 1–3. Later write-ups of the method quote 5–7 ppm of phosphate a week and nobody much minds, which tells you something about how precise any of this needs to be. That distinction matters more than it sounds, and the rest of this page is largely about why.
Why "estimative" is the important word
The name is doing real work, and it is the part that gets dropped. An estimative method is one where being wrong is designed in. You are not measuring uptake, you are not titrating to a target, you are not testing to see whether you got it right. You deliberately overshoot, and then you delete the overshoot before it can matter.
The reason that is worth doing has nothing to do with plants being greedy. It is that a planted tank has a long list of things that can go wrong — light too strong, CO₂ too low, CO₂ unstable, circulation dead, water too soft, iron precipitating, plants melting from a bad transition — and diagnosis is nearly impossible while every one of them is still on the list. EI is a way of crossing one item off permanently. If you are dosing more nitrate than the plants can use and the plants still look bad, then whatever is wrong, it is not nitrate — and you have not had to trust a hobby test kit to know it.
This is the thing to take away, and the reason the method survived twenty years of argument about its numbers. EI is not a better fertilizer than the alternatives; it is an instrument for reducing the number of unknowns in a tank from eight to seven. The people who get the most out of it are the ones who understand that, because they stop reaching for the fertilizer when something looks wrong — the fertilizer is the one variable they have already ruled out.
The water change is the method, not a chore attached to it
Read the forums and you will find people dosing EI amounts and changing water monthly, and describing the result as "EI not working for me". What they have built is not a version of EI. It is the first part of the method without the third, which is simply overdosing.
The mechanism is not complicated: you are adding a fixed amount every week and removing a fraction of what is there. If the fraction is large, the level settles at a predictable ceiling within a few weeks and stays there. If the fraction is small, the ceiling is higher. If there is no water change at all, there is no ceiling — the level rises every week for as long as you keep dosing, and the only thing stopping it is whatever the plants happen to take. Exactly where it settles for a given change size is arithmetic, and the fertilizer calculator works it out for whatever schedule you actually keep.
Which turns the usual question inside out. People ask "do I have to do a 50 % water change for EI?" — as though the change were a tax on the method. It is the other way round: the change is what buys you the freedom not to test, and the dose is sized to it. Change less water and the honest response is to dose less, not to shrug. If large weekly water changes are not something you will actually do — and for plenty of people they are not — that is a completely reasonable position, and it means EI is the wrong method for you rather than that you are doing it badly.
The argument it was built to settle
EI did not arrive into a vacuum. The dominant dry-dosing recipe before it was PMDD — "Poor Man's Dosing Drops", from Paul Sears and Kevin Conlin in 1996 — and its defining feature was that it deliberately left phosphate out. The theory was that algae could be starved: keep phosphate scarce and the algae would be limited before the plants were. A generation of aquarists learned from it that nitrate and phosphate were things to keep low, and that belief is still the single most common thing people arrive at EI already believing.
Barr's counter-position was that this had cause and effect backwards. Starving a tank of phosphate limits the plants first, weak plants lose ground to algae, and the nutrient scarcity you introduced to fix the algae is what fed it. On this account algae is triggered by ammonia and by unstable conditions — a new tank, a disturbed substrate, CO₂ that swings through the photoperiod — rather than by a number on a nitrate kit. Dose generously, keep CO₂ steady, and the algae does not come.
It is hobby-scale, and worth labeling as such. What supports it is two decades of planted tanks running 20–30 ppm of nitrate and several ppm of phosphate with no algae to speak of, which is a genuinely large body of practical evidence and is the reason the argument is now essentially settled in the hobby. What it is not is a controlled study of algae germination against nutrient concentration. "Excess macronutrients do not cause algae in a planted tank" is a claim you can act on with confidence; "algae is caused by ammonia" is a working explanation rather than a demonstrated mechanism, and pages that state it as established fact are overselling it.
What else is in the dose
Macros are three salts and no mystery. The trace mix is where people are vaguest, because it is dosed by its iron and nobody reads the rest of the label. The standard mix, Plantex CSM+B, publishes a full guaranteed analysis, and since a dose is sized to hit 0.2 ppm of iron, everything else in the bag arrives in a fixed ratio to it:
| Element | % of the mix | Per dose | Per week (3 doses) |
|---|---|---|---|
| Iron | 7 % | 200 µg/L | 600 µg/L |
| Manganese | 2 % | 57.1 µg/L | 171.4 µg/L |
| Magnesium | 1.5 % | 42.9 µg/L | 128.6 µg/L |
| Zinc | 0.4 % | 11.4 µg/L | 34.3 µg/L |
| Copper | 0.1 % | 2.9 µg/L | 8.6 µg/L |
| Molybdenum | 0.06 % | 1.7 µg/L | 5.1 µg/L |
| Boron | 0.04 % | 1.1 µg/L | 3.4 µg/L |
Computed by this site's engine from Plantex guaranteed analysis: Fe 7.0 %, Mn 2.0 %, Mg 1.5 %, Zn 0.4 %, Cu 0.1 %, Mo 0.06 %, B 0.04 %, at a full-EI trace dose of 0.2 ppm iron. µg/L is a thousandth of a ppm.
The row people actually want is copper, because shrimp keepers are told — correctly — that copper is lethal to invertebrates, and then left to guess whether their fertilizer contains any. It does: 2.9 µg/L per dose, 8.6 µg/L across a week of full EI, before the water change takes it back out. For scale, a therapeutic copper treatment for parasites runs at 0.5 mg/L — about 58× a full week of EI traces. So the dose is small, it is a plant nutrient rather than a contaminant, and it does not accumulate as long as the water change happens. It is also not zero, which is why cautious shrimp keepers run a copper-free trace mix or dose traces leaner, and why "my fertilizer has no copper in it" is usually wrong.
What a year of it costs
The other thing nobody quantifies, and the reason people stay on bottled fertilizer far longer than they need to: dry salts are absurdly cheap, because you are buying the actual chemical rather than water with a label. A 55 gallon (208 L) tank on full EI, dosed 3 times a week for a year, gets through this:
| Salt | A year uses | A 1 kg bag lasts | A 1 lb bag lasts |
|---|---|---|---|
| Potassium nitrate (KNO₃) | 397 g | 2.5 years | 1.1 years |
| Monopotassium phosphate (KH₂PO₄) | 61 g | 16.5 years | 7.5 years |
| Potassium sulfate (K₂SO₄) | 162 g | 6.2 years | 2.8 years |
| Epsom salt (MgSO₄·7H₂O) | 329 g | 3 years | 1.4 years |
| Plantex CSM+B | 93 g | 10.8 years | 4.9 years |
Computed by this site's engine for 208 L at 7.5 / 1.3 / 7.5 / 1 / 0.2 ppm per dose, 156 doses a year. Potassium sulfate is the top-up after the potassium already in the other two salts, which is why it is smaller than you would expect.
That is 1 kg of salt for the year, which at a middling hobby price of around $20 a kilogram is roughly $21 a year — for a tank that would get through bottled fertilizer at that rate in a couple of months. Half EI is half of it (0.5 kg); the lean PPS-Pro schedule, despite dosing every day, uses 0.4 kg because what matters is the weekly total rather than the number of doses. Buy the smallest bags you can find rather than the cheapest per gram: most of these will outlast your interest in the tank, and potassium nitrate in particular is a strong oxidizer that wants a sealed container and a dry cupboard.
Who it is for — and the case against it
EI was designed for a specific tank: strong light, injected CO₂, a lot of fast-growing stems, and a trimming schedule. In that tank the plants really can outrun a lean dosing scheme, the water change is happening anyway, and the method's forgiveness is worth a great deal. If that is your tank and you are new to dosing, start here — it is the most robust thing to be wrong with.
Where it is the wrong tool:
- Low-tech and low-light tanks. No injected carbon means a fraction of the growth and a fraction of the uptake, so a full dose is just accumulation. Barr's own guidance was a half to a quarter — the calculator scales it.
- Tanks where the water change will not happen. Covered above. This is a scheduling fact about you, not a failing.
- Shrimp-focused tanks, if the copper above bothers you after reading the number — lean dosing and a copper-free trace mix are a reasonable response.
- Tanks whose tap water is nothing like the tank water. A large weekly change is a large weekly swing in GH, KH and temperature. That is a bigger deal than the fertilizer question and it argues for smaller changes and leaner dosing.
And then there is the substantive objection, which comes mostly from the lean-dosing school and deserves a straight answer rather than a dismissal. The argument runs: modern tanks sit on nutrient-rich aquasoil that feeds plants through their roots, many species color up better and grow more compactly at lower nitrate, and dumping 23 ppm a week into the water column is therefore solving a problem the substrate already solved.
Most of that is right. What EI proved — that macronutrient excess does not cause algae — has held up completely. What EI is often taken to imply — that you need the excess — has not: plenty of beautiful tanks run at a third of these numbers, and the aquasoil that makes it possible barely existed when the method was written. The honest summary is that EI is a method for tanks where you want to be certain fertilizer is not the problem, and lean dosing is a method for tanks where you already know it is not. Beginners are usually in the first category, whatever their substrate.
The vocabulary, decoded
Most of the difficulty with EI threads is dialect rather than chemistry.
- Macros — nitrate (NO₃), phosphate (PO₄) and potassium (K). The three you dose in tens of ppm, from three salts.
- Micros / traces / CSM+B — everything in the table above. CSM+B is a brand of chelated trace mix ("Chelated Micronutrients + Boron"), dosed by its iron because iron is the one plants visibly miss.
- Dry salts — the raw chemicals, weighed and dissolved yourself, as opposed to an all-in-one bottle that contains the same things pre-mixed with water.
- Dosing the water column — putting nutrients in the water rather than in the substrate as root tabs. EI is a water-column method by definition.
- ppm — parts per million, which for water is milligrams per liter. "7.5 ppm of nitrate" means 7.5 mg of nitrate ion in every liter of tank water.
- "EI daily" vs "EI 3×" — the same weekly total split into seven small doses or three larger ones. Plants do not notice; you might, because a daily rhythm is easier to remember.
- Lean dosing — the counter-method above: deliberately smaller doses, usually with an aquasoil substrate doing part of the feeding.
- The reset — the weekly water change, named for what it does to the accumulation.
What you'll be told, and what's true
- "EI causes algae — all those nutrients feed it." The single most persistent inheritance from the phosphate-limiting era, and the specific belief the method was written to overturn. Twenty years of tanks dosed this way say otherwise. Algae in an EI tank is nearly always carbon: not the ppm at noon but the swing across the day.
- "With EI you never need a test kit." True of the fertilizer, which is the claim actually being made, and false as general advice. GH and KH still need testing, because calcium and magnesium come from your water rather than the macro bottle, and CO₂ still needs watching. What you stop testing is the thing you are deliberately overdosing.
- "Dry salts are a hassle — bottled ferts are easier." Easier for one evening, at roughly $21 a year against a bottled bill many times that. A scale that reads to 0.01 g costs about the same as two bottles, and dissolving a week of doses into a jug is a five-minute job you do monthly.
- "You must change 50 % every week or it doesn't work." The change size is a dial, not a rule — it sets the ceiling the dose settles at, so smaller changes simply mean smaller doses. What is not negotiable is that some regular change is happening, because without one nothing sets a ceiling at all.
- "Nitrate at 40 ppm will hurt your fish." Higher than the number on a freshwater parameters chart, and the chart is right for the tank it describes — one where nitrate that high means waste is accumulating and something is wrong upstream. In a dosed planted tank the same reading is the method working as designed. Same ion, different diagnosis: what would actually harm the fish in that tank is the CO₂, which is why the injection rate deserves more of your attention than the nitrate does.
Common questions
Do I need test kits to dose EI?
No, and that was the point of it — the original write-up was titled "no need for test kits". The method is built so that being wrong by a third does not matter, because the water change resets the accumulation before it can go anywhere. What is worth testing is not the fertilizer: your GH and KH, because calcium and magnesium come from the water rather than the macro bottle, and your CO2, because that is the variable EI exists to leave you free to blame. A nitrate kit tells you the least of any of them, and hobby phosphate kits at these concentrations are close to decorative.
Can I dose EI without injected CO2?
You can, but you should dose far less of it — the method's own author put low-light tanks at a half to a quarter of full EI. Without injected carbon, plants grow at a fraction of the rate and take a fraction of the nutrients, so a full dose is not feeding faster growth, it is just accumulating until the water change. The scaling is arithmetic rather than judgment, and the fertilizer calculator does it: set the regime to half or type a quarter of each target.
What happens if I miss a dose, or miss the water change?
A missed dose is genuinely nothing — the whole method is built with enough headroom to absorb it, which is what "estimative" buys you. A missed water change is different in kind, because the change is the only thing setting a ceiling on what you have been adding. One skipped week is not a crisis; skipping them as a habit turns EI into plain overdosing, and the level climbs every week with nothing to stop it.
Should I dose on the day of the water change?
Dose after it, not before, and the reason is only that dosing into water you are about to throw away wastes the dose. The conventional rhythm is macros on three days, traces on the days between, and the water change on the seventh, which is a rhythm rather than a requirement — it exists to keep chelated iron and phosphate out of the same bucket of water at the same moment, and to give you a fixed day of the week you will actually remember.