IC50 vs EC50: How to Read Dose-Response Curves in Lab Assays

The two numbers everyone confuses

Open almost any in-vitro pharmacology paper or compound spec sheet and two numbers keep turning up: IC50 and EC50. They're quoted the same way, plotted the same way, and separated by a single letter — yet they answer opposite questions. The information here is for research use only and describes how these values are measured in the laboratory, not how any compound behaves in a person.

For anyone researching a research-grade compound, telling the two apart has a practical payoff: it's the difference between reading a spec sheet critically and mistaking one figure for a verdict. So this article covers what a concentration-response curve actually shows, what IC50 and EC50 each measure, how they differ, how the numbers get fitted, why an IC50 can move between experiments, and how to read a single potency value without asking it to carry more meaning than it can.

What a concentration-response curve actually shows

Before the two numbers make sense, the curve they come from has to. In plain terms, a concentration-response curve graphs "how much compound you add" against "how large the measured response is" — and both IC50 and EC50 are just a single point read off that graph.

The curve almost always takes an S-shape, or sigmoidal form: a flat toe at low concentrations where nothing much changes, a steep climb through the middle where the system is most sensitive, and a plateau up top where adding more stops mattering. That upper plateau is the maximal response, often written as Emax.

The x-axis is usually plotted on a logarithmic concentration scale. Compressing a range that can span nanomolar to millimolar into a readable line is the main reason, though it carries a caveat: a log axis can make it look as if there's a hard threshold concentration when biologically there's none. Three landmarks are worth holding onto — the baseline, the steep midpoint, and the plateau. Both IC50 and EC50 name that midpoint, for two different kinds of experiment.

IC50 — the number for inhibition assays

Short version: IC50 tells you how little compound it takes to knock a signal down by half. Formally, the half maximal inhibitory concentration is the concentration that inhibits a biological process, enzyme, or signal by 50% in vitro.

It's the standard summary for experiments where the compound turns something off: antagonist assays, competition-binding assays, and cell-viability readouts like the MTT assay, where researchers watch a signal fall as compound concentration rises. In a competition-binding format specifically, the IC50 is the concentration that displaces half of a labeled reference ligand from its target.

Reading the number is straightforward once the direction is clear. A lower IC50 means less compound is needed to reach 50% inhibition, which is described as higher potency. Because the useful values span many orders of magnitude, they're often reported logarithmically as pIC50, where pIC50 = −log10(IC50); a higher pIC50 means a more potent inhibitor. None of this, on its own, says anything about how the compound behaves outside the assay. It describes the experiment, nothing more.

EC50 — the number for stimulation assays

Short version: EC50 tells you how little compound it takes to produce half of the effect that compound can produce. The half maximal effective concentration is the concentration that generates a response halfway between the baseline and the compound's own maximum (Emax), after a specified exposure time.

EC50 is the standard summary for experiments where the compound turns something on: agonist and stimulator assays. That includes functional agonist readouts such as calcium flux assays and GPCR signaling readouts like cAMP accumulation, where the measured response climbs as concentration rises. As with IC50, a lower EC50 means higher potency, and it's often written as pEC50 = −log10(EC50).

One distinction is worth carrying carefully, because it's where a lot of misreadings begin. Potency is not the same as affinity, and neither is the same as efficacy. Affinity — captured by the dissociation constant Kd — describes how well a compound binds its target. Efficacy describes how large a response it can drive once bound. A compound's observed potency reflects both working together, which is why an EC50 alone can't tell you whether a compound binds tightly, drives a big response, or some mix of the two.

IC50 vs EC50: the difference in one view

Here's the whole distinction in a sentence: IC50 summarizes an experiment where the compound turns a response down, and EC50 summarizes one where the compound turns a response up. Same 50% midpoint idea, opposite experiments.

That's why the choice of metric follows the assay design rather than the compound. An antagonist or inhibition experiment yields an IC50; an agonist or stimulation experiment yields an EC50, as the Assay Guidance Manual lays out for structure-activity work. It also explains why you can't convert one into the other with a formula — they're built from different experiments with different reference points. The two get confused mainly because they share the "50% concentration" language and sit at the same place on a sigmoidal curve, but the experiments underneath them point in opposite directions.

How the number gets fitted — the four-parameter logistic model

Here's a point worth internalizing: an IC50 or EC50 is almost never measured directly. It's fitted from the shape of the entire curve.

The standard tool is the four-parameter logistic (4PL), or Hill-slope, model, and the name is the checklist. The four parameters are the top asymptote, the bottom asymptote, the Hill slope, and the midpoint — and that midpoint is the potency value being reported. Most papers quote a "relative" IC50, meaning the concentration halfway between the fitted top and bottom of that specific curve, rather than an absolute 50% of some fixed scale.

The Assay Guidance Manual flags one quality rule worth knowing as a reader: a trustworthy IC50 or EC50 has measured data points on both sides of the midpoint. If every point sits above or below the 50% mark, the value is an extrapolation, not an interpolation, and it should be reported as a bound — something like "<Xmin" or ">Xmax" — rather than a hard number.

The Hill slope is information in its own right, not just curve-fitting bookkeeping. It describes how steeply the curve rises through its middle: a slope near 1 is consistent with simple one-to-one binding, while a slope above 1 points to positive cooperativity. Two compounds can share the same midpoint and still behave very differently across the concentration range, and the slope is where that difference shows up.

Why an IC50 shifts between experiments

The plain-English lead here matters, because it heads off a common misreading. When the same compound gives two different IC50 values in two labs, that's usually expected behavior — not a mistake.

IC50 is condition-dependent. It moves with the concentration of the competing agonist or substrate, the incubation time, the number of cells in the well, and the assay format itself. Work analyzing dozens of compounds across many cell lines has shown these systematic variations are real and predictable rather than random noise. So two groups running the same compound under different conditions can legitimately report different IC50s.

This is where an affinity constant earns its keep. Unlike IC50, the inhibition constant Ki is an absolute value, independent of the agonist concentration used in the experiment. The Cheng-Prusoff equation converts an IC50 into a Ki, giving good estimates at high agonist concentrations and less reliable ones at low concentrations. When an absolute, condition-independent comparison is needed, Ki is the anchor — which is exactly what radioligand binding assays that report Kd and Ki are built to provide.

Reading the number responsibly

The honest summary: a single IC50 or EC50 compresses a whole experiment, and compression loses detail. Treating one figure as the last word is the most common way these numbers get misused.

The cancer-drug analysis cited above makes the point directly — focusing on potency alone can obscure how a compound actually behaves. Metrics like the maximum effect (Emax), the Hill slope, the area under the curve, and growth-normalized values (GI50) each capture something a midpoint misses: how big the effect gets, how steep the transition is, and how the whole curve integrates. The same work found that different parameters carry the most weight at different parts of the curve — EC50 tends to dominate at low concentrations, IC50 in the middle, and Emax at the top.

For anyone reading a spec sheet or a paper, the takeaway we'd offer is to treat a potency number as one input among several, framed as what was observed in a defined in-vitro experiment. Paired with affinity (Kd or Ki) and efficacy (Emax), an IC50 or EC50 becomes genuinely informative. On its own, it's a single coordinate on a curve that has a lot more to say.

Frequently Asked Questions

Is a lower IC50 or EC50 better?

A lower value means the compound reaches its half-maximal effect at a smaller concentration, which is described as higher potency. It is not a value judgment about quality — potency is only one axis. A compound with a low IC50 but a shallow curve or a small maximum effect may be a weaker overall inhibitor than the single number suggests, so potency should be read alongside efficacy and curve shape.

Can you convert an IC50 to an EC50?

Not with a simple formula. They summarize different experiments — IC50 comes from an assay measuring inhibition of a response, EC50 from an assay measuring a compound producing a response. You can convert an IC50 to Ki, a condition-independent affinity value, using the Cheng-Prusoff equation, but that is a different quantity from EC50.

Why did the same compound give two different IC50 values in two labs?

IC50 is condition-dependent. It shifts with agonist or substrate concentration, incubation time, cell number, and assay format. Two labs using different conditions can report different IC50s for the same compound without either being wrong. This is why affinity constants like Ki, which are independent of those conditions, are used when an absolute comparison is needed.

What is the Hill slope on a concentration-response curve?

The Hill slope, or Hill coefficient, describes how steeply the curve rises through its middle section. A slope near 1 is consistent with simple one-to-one binding; a slope above 1 suggests positive cooperativity; a shallow slope below 1 can point to more complex behavior. It is reported alongside the IC50 or EC50 because two compounds can share a midpoint yet behave very differently across the concentration range.

The Bottom Line

IC50 and EC50 are the same idea wearing two hats: a 50% midpoint read off a sigmoidal curve, applied to opposite experiments — inhibition for IC50, stimulation for EC50. Once the direction is clear, the rest follows. The number is a fitted summary, it depends on the conditions of the experiment, and it deliberately hides the curve's shape and the size of the maximum effect. Read it next to an affinity value (Kd or Ki) and an efficacy value (Emax), and it becomes one solid data point among several. For readers working through the related assay explainers, that habit — treating potency as an input, not an outcome — is what keeps a spec sheet honest.

For research use only. Not for human or animal consumption of any kind. The information in this article is for educational purposes only and is not intended to diagnose, treat, cure, or prevent any disease. The statements made have not been evaluated by the U.S. Food and Drug Administration. These products are NOT FDA APPROVED. Please consult with a licensed healthcare professional before making any decisions regarding your health or research.

Optides LLC is a chemical supplier. Optides LLC is not a compounding pharmacy or chemical compounding facility as defined under 503A of the Federal Food, Drug, and Cosmetic Act. Optides LLC is not an outsourcing facility as defined under 503B of the Federal Food, Drug, and Cosmetic Act.