Connect via MCP →

Enter Calculation

Formula

Advertisement

Results

Z-Score for the 97.5th Percentile
1.96
standard deviations from the mean
Percentile 97.5%
Z-Score 1.96

What Is a Percentile-to-Z-Score Calculator?

This tool converts a percentile into its corresponding z-score on the standard normal distribution (mean 0, standard deviation 1). A z-score tells you how many standard deviations a value lies above or below the mean. Because percentiles describe the share of data below a point, finding the z-score requires the inverse of the normal cumulative distribution function, written \(\Phi^{-1}\).

Standard normal bell curve with shaded left area p and a vertical line at z marking the percentile boundary
A percentile p is the shaded area to the left of the z-score under the standard normal curve.

How to Use It

Enter a percentile between 0 and 100 — for example, 90 means "90% of values fall below this point." The calculator returns the z-score. Percentiles below 50 produce negative z-scores (below the mean), exactly 50 gives 0, and values above 50 give positive z-scores.

The Formula Explained

If \(p\) is the percentile divided by 100, then

$$z = \Phi^{-1}(p)$$

where \(\Phi\) is the standard normal CDF. There is no simple closed form for \(\Phi^{-1}\), so this calculator uses Acklam's rational approximation, which is accurate to roughly \(1\times10^{-9}\) across the full range.

Advertisement
Diagram showing the inverse relationship: percentile p maps through inverse normal CDF to z-score on the axis
The inverse normal CDF turns a cumulative probability p into its z-score.

Worked Example

Suppose you want the z-score for the 97.5th percentile. Set \(p = 0.975\). The inverse normal CDF returns

$$z = \Phi^{-1}(0.975) \approx 1.9600$$

This is the familiar critical value used for a 95% confidence interval (since 2.5% lies in each tail).

FAQ

What z-score equals the 50th percentile? Exactly 0, since the median of a normal distribution sits at the mean.

Why can't I enter 0 or 100? The z-scores for 0 and 100 are negative and positive infinity. The calculator clamps extreme inputs to return a very large finite value instead.

Is this for a standard normal distribution? Yes. To convert to a real distribution with mean \(\mu\) and standard deviation \(\sigma\), use \(x = \mu + z\cdot\sigma\).

Last updated: