What Is a Z-Score and How to Interpret It

A z-score measures how many standard deviations a single value sits above or below the mean of a dataset. A z-score of +2 means a value is two standard deviations above average. A z-score of −1.5 means it is one and a half standard deviations below. That single number lets you compare values from completely different datasets on the same scale.

Z-scores are used everywhere from standardized testing to medical diagnostics to financial risk analysis. When a doctor reports that a child's bone density z-score is −2.5, that number has a precise statistical meaning — it places the child in the lowest 0.6% of the reference population. Understanding what is a z-score and how to interpret it is not just a classroom exercise; it is how professionals contextualize data every day.

Use the free Z-Score Calculator at RoughTools to convert any value to a z-score and percentile instantly — or follow the step-by-step method below.

The Z-Score Formula

The z-score formula is one of the cleanest in statistics — three values, one result.

z = (x − μ) / σ

Where:

x — the individual data point you are converting to a z-score
μ — the mean (average) of the population or reference group
σ — the standard deviation of the population, which measures how spread out the values are
z — the resulting z-score: how many standard deviations x sits from the mean

A positive z-score means the value is above average. A negative z-score means it is below average. A z-score of zero means the value equals the mean exactly.

Worked example: student exam score

A student scores 87 on a chemistry exam. The class mean is 71.3 and the standard deviation is 9.4.

Step 1 — Identify the values:

x  = 87    (the student's score)
μ  = 71.3  (class mean)
σ  = 9.4   (class standard deviation)

Step 2 — Subtract the mean from the value:

x − μ = 87 − 71.3 = 15.7

Step 3 — Divide by the standard deviation:

z = 15.7 / 9.4
z = 1.67

The result: a z-score of +1.67 means this student scored 1.67 standard deviations above the class average. Looking that up in a standard normal distribution table, a z-score of 1.67 corresponds to approximately the 95th percentile — meaning this student scored higher than roughly 95% of the class. That context is what makes z-scores valuable; "87 out of 100" tells you less than "95th percentile of this specific class."

How to Calculate Z-Score Step by Step

Confirm you have three values: the individual score, the group mean, and the group standard deviation. If you are working with a sample rather than a full population, use the sample mean (x̄) and sample standard deviation (s) in place of μ and σ. The formula is identical — the distinction only matters when labeling your result for a statistics course or report.
Subtract the mean from the individual value. The result is the raw deviation — how far the value sits from the center. This number can be positive, negative, or zero. A score of 71.3 in the chemistry example would give a deviation of exactly 0, because it equals the mean.
Divide the deviation by the standard deviation. This converts the raw deviation into standard deviation units. The result is your z-score. If the standard deviation is 9.4 and the deviation is 9.4, the z-score is exactly 1.0 — the value is precisely one standard deviation above the mean.
Interpret the sign and magnitude. The sign (+ or −) tells you the direction. The magnitude tells you how extreme. A z-score between −1 and +1 is within one standard deviation of the mean — where about 68% of normally distributed values fall. A z-score beyond ±2 is in the outer 5% of the distribution. Beyond ±3 is exceptionally rare — less than 0.3% of values.
Convert to a percentile if needed. Use a z-table (standard normal distribution table) or the z-score calculator to find the percentage of values below your z-score. This is called the cumulative probability and is the most intuitive way to communicate a z-score to a non-statistics audience.
Verify the result is directionally correct. If the original value is above the mean, the z-score must be positive. If below, it must be negative. A z-score whose sign contradicts the direction of the deviation is always an arithmetic error — recheck which value you subtracted from which.

Pro tip: When comparing z-scores across two different datasets, the z-scores are directly comparable only if both underlying distributions are approximately normal. Comparing a z-score from a normal distribution to one from a heavily skewed distribution can be misleading — the percentile conversion assumes normality.

How Do You Convert a Z-Score to a Percentile?

A z-score converts to a percentile by looking up the cumulative area under the standard normal curve to the left of that z-score. This is the percentage of values in a normal distribution that fall below your z-score.

The most common z-scores and their percentile equivalents:

| Z-Score | Percentile | Plain meaning | |---|---|---| | −3.0 | 0.13% | Bottom 0.13% of the distribution | | −2.0 | 2.28% | Lower than 97.7% of values | | −1.0 | 15.87% | Lower than 84.1% of values | | 0.0 | 50.00% | Exactly at the median | | +1.0 | 84.13% | Higher than 84.1% of values | | +1.67 | 95.25% | Higher than 95.2% of values | | +2.0 | 97.72% | Higher than 97.7% of values | | +3.0 | 99.87% | Top 0.13% of the distribution |

For the chemistry exam example (z = 1.67), the percentile is approximately 95.3%. That means the student scored better than about 95 out of every 100 students in the same group.

One detail that catches people off guard: the percentile from a z-table is always the percentage of values below your value, not the percentage at or above. A student at the 95th percentile scored higher than 95% of peers, not equal to the top 5%. If you need the probability of a value above your score, subtract the table value from 1 (or 100%). For z = 1.67: 1 − 0.9525 = 0.0475, or 4.75% of students scored higher.

What Is a Good Z-Score?

"Good" depends entirely on the context — whether a high or low z-score is desirable varies by what is being measured.

For test scores, grades, or performance ratings, a positive z-score is better. A z-score of +1.5 on a standardized exam places you in roughly the 93rd percentile — strong performance.

For medical risk markers, a negative or near-zero z-score is often preferred. A z-score of −0.3 on a blood pressure measure means your reading is slightly below the population average, which may indicate lower cardiovascular risk. A z-score of +2.8 on a tumor marker test would be a significant warning flag.

For quality control in manufacturing, a z-score close to zero is ideal — it means output is near the target specification. Scores beyond ±3 trigger defect review.

A general-purpose interpretation framework:

|z| < 1.0 — within one standard deviation of the mean; typical range
1.0 ≤ |z| < 2.0 — moderately above or below average; notable
2.0 ≤ |z| < 3.0 — significantly above or below; in the outer 5% of a normal distribution
|z| ≥ 3.0 — extreme; less than 0.3% of normally distributed values fall here; often flagged as an outlier

In IQ testing, which is standardized with a mean of 100 and SD of 15, a score of 130 yields z = (130 − 100) / 15 = +2.0, placing it at approximately the 97.7th percentile. Scores at or above z = +2.0 are the threshold for "gifted" classification in many frameworks.

What Is the Difference Between a Z-Score and a T-Score?

A z-score and a t-score both measure how far a value sits from a mean in standard deviation units, but they are used in different situations and have different scales.

Z-scores assume you know the true population standard deviation (σ). T-scores are used when the population standard deviation is unknown and must be estimated from a sample. The t-distribution has heavier tails than the normal distribution — especially with small samples — which accounts for the additional uncertainty.

The second difference is scale. In inferential statistics, z-scores and t-scores use roughly the same scale (both centered at 0). But in educational and psychological testing, t-scores are rescaled to a mean of 50 and standard deviation of 10 to avoid negative numbers. On this scale, a t-score of 60 is one standard deviation above average — equivalent to a z-score of +1.0.

| Feature | Z-Score | T-Score (testing context) | |---|---|---| | Center | 0 | 50 | | Standard deviation unit | 1 | 10 | | Negative values | Yes | Rare (designed to avoid them) | | Used when | Population SD is known | Population SD unknown or in psychometrics | | Example scale | SAT score normalization | Personality assessments, bone density |

The practical rule: if you are working with a full population or a large sample (n > 30) and know the standard deviation, use z-scores. If you have a small sample and an estimated standard deviation, use the t-distribution. The confidence interval calculator automatically applies the correct distribution based on your sample size.

Common Mistakes to Avoid When Using Z-Scores

Subtracting x from μ instead of μ from x. The formula is (x − μ) / σ, not (μ − x) / σ. Reversing the subtraction flips the sign of the z-score. A student who scored above average would get a negative z-score — the wrong direction. Always put the individual value first.
Using the sample standard deviation when the population SD is required. Z-scores technically use the population standard deviation σ. If you use a sample SD (s) when σ is not known, the result is technically a t-statistic, not a z-score. For large samples (n > 30) the difference is negligible, but for small samples it matters. Clarify which type of SD you have before calculating.
Treating z-scores as percentiles without the conversion step. A z-score of 1.67 is not a percentile of 1.67 or 167. It is a standardized deviation score that requires a z-table or normal distribution calculation to convert to a percentile. Confusing these two is common and leads to completely wrong interpretations.
Applying z-scores to non-normal data and assuming the percentile table holds. The z-to-percentile table assumes a normal distribution. For heavily skewed data — incomes, social media follower counts, house prices — the percentile implied by a z-score may be significantly off. Always check the distribution shape before converting z-scores to percentiles. The statistics calculator shows skewness so you can assess this.
Forgetting that z-scores only work within a defined reference group. A z-score of +2.1 means very different things depending on the reference population. A z-score of +2.1 for height among 10-year-old girls is not the same as +2.1 among adult men. Always state clearly what population the mean and SD come from — the z-score is only meaningful relative to that group.

Frequently Asked Questions

What does a z-score of 0 mean? A z-score of 0 means the value equals the mean of the reference group exactly — it is neither above nor below average. In a normal distribution, the mean sits at the 50th percentile. A z-score of 0 is not "bad" or unusual — it simply means the value is exactly typical for that group.

What if my z-score is greater than 3 or less than −3? A z-score beyond ±3 is statistically unusual — in a normal distribution, less than 0.3% of values fall outside this range. It may indicate an outlier, a measurement error, or a genuinely extreme value worth investigating. In quality control, values beyond ±3 standard deviations trigger review. In medical contexts, a z-score below −2.5 (like bone density) often crosses a clinical threshold. The z-score is still mathematically valid; the interpretation depends on whether the extreme value is plausible for your measurement.

What is the difference between a z-score and a standard score? They are the same thing. "Standard score" is the broader term; z-score is the most common type of standard score. Both refer to the result of the formula (x − μ) / σ. Other standardized score systems (like IQ scores or SAT scores) are transformations of z-scores — they rescale the distribution to a different mean and SD while preserving the same rank ordering.

How many standard deviations from the mean is unusual? By convention, a value more than 2 standard deviations from the mean (|z| > 2) is considered statistically unusual — it falls in the outer 5% of a normal distribution. A value more than 3 standard deviations away (|z| > 3) is considered a statistical outlier — roughly 1 in 370 values falls here by chance alone. These thresholds are conventions, not absolute rules, and the appropriate cutoff depends on the specific application.

When should I use z-scores instead of raw scores? Use z-scores when you need to compare values across different scales or units. For example, comparing a student's performance on a math test (mean 68, SD 12) and an English test (mean 81, SD 7) is misleading with raw scores — a student who scored 80 on math and 88 on English did better relative to peers on math (z = 1.0) than on English (z = 1.0). Use the z-score calculator when you are normalizing data before analysis, ranking performance across different groups, or interpreting standardized test results.

Use the Free Z-Score Calculator

The Free Z-Score Calculator at RoughTools converts any value to a z-score and its corresponding percentile in one step — just enter the individual value, the group mean, and the standard deviation. It shows the z-score, the cumulative percentile, and the probability of values above and below your result, so you get the full statistical picture without a z-table. No account needed, no data stored, completely free.

Free Z-Score Calculator →

You might also need:

Standard Deviation Calculator — calculate the SD you need before computing z-scores
Confidence Interval Calculator — build confidence intervals using z-scores and sample data
Statistics Calculator — get mean, SD, skewness, and full descriptive statistics for your dataset
Mean Median Mode Calculator — find the mean you need as an input to the z-score formula