Z-Score Calculator

Calculate a z-score — how many standard deviations a value or a sample mean sits from the mean — with the 68-95-99.7 empirical rule and step-by-step working.

Loading calculator…

What a z-score measures

A z-score (or standard score) converts a raw value into units of "how many standard deviations from the mean." It answers one question in a form that works the same way no matter the original units: is this value typical, or unusual — and by how much?

  • z = 0 — exactly at the mean.
  • z > 0 — above the mean; the larger z is, the further above.
  • z < 0 — below the mean; the more negative z is, the further below.

Because z-scores strip away the original units, they let you compare a test score, a height, a temperature, and a stock return on the same scale — all as "number of standard deviations from typical."

Two formulas, one idea

This calculator has two modes:

Single point — how unusual is one observation?

z=xμσz = \frac{x - \mu}{\sigma}

Sample mean — how unusual is a sample's average?

z=xˉμσ/nz = \frac{\bar{x} - \mu}{\sigma / \sqrt{n}}

The sample-mean version divides by the standard error (σ/√n) instead of σ directly. Averaging many observations cancels out noise, so sample means cluster more tightly around μ than individual values do — the larger the sample size n, the tighter that clustering, and the smaller the standard error becomes. Using σ/√n (rather than σ) correctly reflects that a sample mean of 16 observations is a much steadier number than any one of those 16 observations on its own.

Tip: Same gap from the mean, different z: a single test score 2 points above average might be unremarkable (small z), but a class average 2 points above the historical average can be a large z — because sample means don't wander around as much as individual scores do.

Worked example: single point

Using the same 7-value data set from the standard deviation calculator42, 54, 65, 47, 59, 40, 53 — with mean μ ≈ 51.4286 and population standard deviation σ ≈ 8.364:

For x = 65:

  1. z = (65 − 51.4286) / 8.364 = 13.5714 / 8.364
  2. z ≈ 1.62
  3. |z| = 1.62 ≤ 2, so this value falls in the 95% empirical-rule band — moderately above average, but not unusual.

For x = 40 (the smallest value in the set):

  1. z = (40 − 51.4286) / 8.364 = −11.4286 / 8.364
  2. z ≈ −1.37 — negative, because 40 is below the mean.

Worked example: sample mean

A sample of n = 16 observations has a mean x̄ = 52, drawn from a population with μ = 50 and σ = 8:

  1. Standard error = σ/√n = 8/√16 = 8/4 = 2.
  2. z = (x̄ − μ)/(σ/√n) = (52 − 50)/2 = 2/2 = z = 1.

Compare that to a single observation of x = 52 from the same population using the single-point formula: z = (52 − 50)/8 = 0.25 — a much smaller z, because one observation is far noisier than the average of 16.

The empirical rule: 68-95-99.7

For data that's approximately normally distributed (bell-curve shaped), the share of values within each number of standard deviations of the mean follows a predictable pattern:

RangeShare of the dataThis calculator's band
|z| ≤ 1≈ 68%68
|z| ≤ 2≈ 95%95
|z| ≤ 3≈ 99.7%99.7
|z| > 3< 0.3%beyond — rare, worth a second look as a possible outlier

This calculator classifies your result into one of these four bands automatically, alongside the exact z value.

Tip: Note what this calculator deliberately does not do: it has no built-in z-to-probability lookup table (the kind of table that answers "what exact percentage of a normal distribution is below z = 1.62?"). That's a separate, more precise tool — this one is built for the fast, everyday question of which broad empirical-rule band a value falls in.

Where z-scores show up

Z-scores are the backbone of standardized testing (comparing scores across different test forms or years), quality control (flagging measurements too many standard deviations from a target), finance (assessing how unusual a return or price move is), and outlier detection generally. For the mean and standard deviation this calculation is built on, see mean, median & mode and standard deviation; for another way to compare a value against a whole data set directly, see percentile.

Frequently asked questions

What does a z-score actually mean?
A z-score is a value converted into "how many standard deviations from the mean" units. A z-score of 0 means the value equals the mean; +1 means it's one standard deviation above the mean; −2 means it's two standard deviations below. Because z-scores are unitless, they let you compare values from completely different scales — a test score, a height, a stock return — on the same footing.
What's the difference between the single-point and sample-mean modes?
Single point asks "how unusual is this one observation?" — z = (x − μ)/σ, comparing one value directly to the population's spread. Sample mean asks "how unusual is this sample's average?" — z = (x̄ − μ)/(σ/√n), comparing a sample mean to the population mean using the standard error (σ/√n) instead of σ directly. Sample means cluster more tightly around μ as n grows, so the same gap from μ produces a larger |z| in sample-mean mode than in single-point mode.
Why does the sample-mean formula divide by σ/√n instead of σ?
σ/√n is the standard error of the mean — the expected spread of sample means around the true population mean μ, not the spread of individual values. As sample size n grows, averaging cancels out more of the random noise, so sample means bunch up closer to μ and the standard error shrinks. Dividing by this smaller number (instead of σ) correctly reflects that a sample mean is a much more precise estimate than any single observation.
What is the 68-95-99.7 (empirical) rule?
For data that's approximately normally distributed, about 68% of values fall within 1 standard deviation of the mean (|z| ≤ 1), about 95% fall within 2 (|z| ≤ 2), and about 99.7% fall within 3 (|z| ≤ 3). This calculator reports which of those three bands your z-score falls in, plus a fourth "beyond 3" band for the rare, more-extreme-than-usual cases.
Does this calculator give an exact probability for my z-score?
No — deliberately not. Exact z-to-probability conversion requires a full standard normal table (or a cumulative distribution function), which is a separate, more detailed tool. This calculator answers a simpler, faster question — which of the three empirical-rule bands (68%, 95%, 99.7%) your value falls into — without implying more precision than a quick mental-math check needs.
Can a z-score be negative, and what does that mean?
Yes — a negative z-score simply means the value is below the mean rather than above it. z = −1.5 is exactly as far from the mean as z = +1.5, just on the other side. The sign tells you direction (above/below); the absolute value |z| tells you how far, which is what the empirical-rule bands use.
Why must σ be nonzero?
σ measures spread, and z-score is defined as how many "spread units" a value sits from the mean. If σ = 0, every value in the population is identical to μ, so there's no spread to measure against — the division (x − μ)/σ is undefined, the same way any division by zero is.

Related calculators