What does a z-score tell you?

A z-score tells you how far a value is from the mean, measured in standard deviation units. A z-score of 0 means the value equals the mean. A z-score of +1 means one standard deviation above the mean; −2 means two standard deviations below. Under a normal distribution: ~68% of values have z-scores between −1 and +1; ~95% between −2 and +2; ~99.7% between −3 and +3. This is the empirical rule (68–95–99.7 rule).

How do you interpret a z-score in practice?

Example: a student scores 82 on an exam with mean 70 and standard deviation 10. z = (82 − 70) / 10 = 1.2. This means the student scored 1.2 standard deviations above the class average — approximately the 88th percentile. Z-scores allow comparisons across different exams, scales, or populations without needing to know the original units.

What is the difference between z-score and t-score?

Both standardise a value relative to a distribution. Use a z-score when the population standard deviation (σ) is known, or when the sample size is large (n ≥ 30). Use a t-score when the population standard deviation is unknown and must be estimated from the sample — typically with small samples. For large samples, z and t values converge to nearly the same result.

How do you convert a z-score to a percentile?

A percentile represents the percentage of values at or below a given point. For a normal distribution, a z-score of 0 corresponds to the 50th percentile. A z-score of +1 corresponds to approximately the 84.1st percentile. A z-score of −1.645 corresponds to the 5th percentile (often used as the critical value for one-tailed 5% significance tests). This calculator uses the normal CDF to convert z-scores to percentiles automatically.

Can z-scores be used with non-normal distributions?

Yes — the z-score formula itself works for any distribution. However, the percentile and probability interpretations (using the normal CDF) are only accurate for data that is approximately normally distributed. For heavily skewed, bimodal, or non-normal distributions, z-scores still measure distance from the mean in SD units, but the corresponding percentile estimates will be inaccurate.

What are typical z-score critical values?

Common critical values: ±1.645 for 90% confidence interval (5% in each tail); ±1.960 for 95% CI (2.5% each tail); ±2.326 for 98% CI; ±2.576 for 99% CI. In one-tailed hypothesis tests: 1.282 (10%), 1.645 (5%), 2.326 (1%). These values come from the standard normal table and are used extensively in statistical inference.

Z-Score Calculator

Name: Z-Score Calculator — Standardise Values & Find Percentiles
Availability: OnlineOnly
Author: CalcMulti Editorial Team

Reviewed by CalcMulti Editorial Team·Last updated: February 2026·← Statistics Hub

A z-score (standard score) expresses how many standard deviations a value lies above or below the mean of its distribution. It lets you compare values from different datasets on a common scale — making it one of the most widely used tools in statistics.

This calculator computes the z-score from a raw value, finds the corresponding percentile rank, and calculates cumulative normal distribution probabilities.

Formula

z = (x − μ) / σ

z: z-score (standard score)
x: the raw value you are standardising
μ: mean of the distribution
σ: standard deviation of the distribution (σ > 0)

Enter a raw value plus the distribution's mean and standard deviation to compute the z-score and percentile.

Value (x)

Mean (μ)

Std Dev (σ)

The 68–95–99.7 Rule (Empirical Rule)

|z| ≤ 1

68.27%

Within 1 std dev of mean

|z| ≤ 2

95.45%

Within 2 std devs of mean

|z| ≤ 3

99.73%

Within 3 std devs of mean

Under a normal distribution, about 68% of observations fall within one standard deviation of the mean, 95% within two, and 99.7% within three. Values with |z| > 3 are rare — less than 0.3% of the distribution.

Common Critical Z-Values

Z-Score	P(X ≤ z)	Percentile	Use Case
−2.576	0.50%	0.5th	99% CI lower tail
−1.960	2.50%	2.5th	95% CI lower tail
−1.645	5.00%	5th	90% CI lower tail / one-tail 5%
0.000	50.00%	50th	Mean — median for symmetric dist.
+1.282	90.00%	90th	One-tail 10% significance
+1.645	95.00%	95th	One-tail 5% / 90% CI upper tail
+1.960	97.50%	97.5th	95% CI upper tail
+2.326	99.00%	99th	98% CI upper tail / one-tail 1%
+2.576	99.50%	99.5th	99% CI upper tail
+3.000	99.87%	99.87th	Six-sigma quality baseline

Z-Score vs T-Score — When to Use Which

Condition	Use Z	Use T
Population σ known	✓
Population σ unknown		✓
Large sample (n ≥ 30)	✓	Acceptable
Small sample (n < 30)	Not ideal	✓
Data is normal	✓	✓
Comparing means across populations	✓ (known σ)	✓ (unknown σ)

As sample size grows, the t-distribution approaches the standard normal distribution, so z and t values converge for n ≥ 30.

Common Mistakes with Z-Scores

Using z-scores on non-normal data

Z-scores and percentile lookups assume a normal distribution. Applying them to heavily skewed data (income, web traffic) gives misleading percentiles. Check your distribution first with a histogram or Q-Q plot.

Forgetting to use the correct standard deviation

Use population SD (σ) when standardising against a known distribution. Use sample SD (s) when working with a dataset sample. The formula looks the same but the interpretation — and the error it introduces — differs.

Reading z-score as probability directly

A z-score of 1.5 is not a 1.5% probability. You must look up Φ(1.5) ≈ 0.9332, meaning 93.3% of values fall below. Use the cumulative distribution function, not the z-score value itself.

Mixing one-tailed and two-tailed areas

P(Z < 1.96) = 97.5%, but for a 95% confidence interval you need both tails, so you use z = ±1.96 (2.5% in each tail). Confusing one-tailed and two-tailed interpretations is one of the most common errors in hypothesis testing.

Case Study: Detecting Out-of-Spec Tablets in Pharmaceutical QC

A quality control engineer at a pharmaceutical plant monitors tablet weight on the production line. The target weight is μ = 500 mg with a process standard deviation of σ = 4.2 mg, established from a large historical population of production runs.

During a routine sample of 80 tablets, one weighs 493.1 mg. The z-score is (493.1 − 500) / 4.2 = −1.64. Looking up Φ(−1.64) ≈ 0.0505 — just over 5% of tablets from this distribution are expected to fall this low or lower.

The action limit is z < −2.0 (below 2.3% of the distribution). At z = −1.64, the tablet is within spec but is flagged for trend monitoring. If three consecutive tablets show z < −1.5, it triggers a process review — catching variability drift before it becomes a compliance failure.

Related Calculators

Normal Distribution Calculator

Convert z-score to probability P(X < x)

T-Test Calculator

Hypothesis test using t-statistic

Standard Deviation Calculator

Compute σ for your dataset first

Mean Calculator

Compute μ for your dataset

Chi-Square Calculator

Categorical data distribution test

Correlation Calculator

Pearson r for two variables

Confidence Interval Calculator

Uses z-scores for CI construction

Z-Table (Standard Normal)

Cumulative probabilities for z-scores −3.4 to 3.4

Z-Score Formula Guide

Formula derivation, interpretation & examples

Z-Score vs Percentile

How z-scores map to percentile ranks

T-Test vs Z-Test

Which test to use and when

Statistics Hub

All statistics calculators & guides

Disclaimer

This calculator is for educational purposes only and does not constitute professional advice. Results are based on standard mathematical formulas. Always verify critical calculations with a qualified professional before making important decisions.