Standard Deviation Graph in Excel: Create and Read It

Charts & pivot tables · Updated November 18, 2026

Excel does not have a single chart type called a standard deviation graph. To show a distribution, calculate the mean and standard deviation, then plot a normal curve or make a histogram; to show variation around plotted averages, add standard deviation error bars. The right choice depends on whether you want to explain a set of observations or compare categories.

What a standard deviation graph shows

Standard deviation describes how spread out numerical values are around their mean. A small standard deviation means observations tend to cluster near the average; a larger one means they are more dispersed. A graph can make that spread easier to see, but the phrase standard deviation graph can refer to several different displays.

A bell-shaped curve is often used to illustrate a normal distribution. Its center is the mean, and its width is determined by the standard deviation. For a roughly normal distribution, about 68% of observations fall within one standard deviation of the mean, about 95% within two, and about 99.7% within three. Those are approximate properties of a normal distribution, not guarantees for every dataset.

A histogram, by contrast, shows the observed frequency of values in intervals. Error bars can show variation around points or columns in a chart. These displays answer different questions: a curve describes a theoretical distribution, a histogram summarizes the observed data, and error bars show uncertainty or spread for plotted results.

Choose the right chart for your data

  • Use a histogram to see how a single set of measurements is distributed, including whether it is skewed or has multiple clusters.
  • Use a normal curve when you want to show the bell-shaped distribution implied by a mean and standard deviation, or compare that model with data.
  • Use error bars when a chart compares group means and you want to display variability for each group.
  • Use a box-and-whisker chart when comparing medians, quartiles, and possible outliers is more useful than plotting a standard deviation.

Do not assume that every dataset follows a normal distribution. A bell curve calculated from the mean and standard deviation is a model; it does not reproduce the actual shape of your observations. Inspect a histogram before relying on a normal curve, especially when there are few values, strong skew, or extreme outliers.

Prepare the data and calculate standard deviation

Put one numerical observation in each cell of a single column, with an optional heading above the values. For the worked example, enter 8, 9, 10, 10, 11, 11, 12, 12, 13, and 15 in H2:H11. The mean is 11.1. The sample standard deviation is about 2.02, so the data typically vary by roughly two units around the mean.

With the ten sample observations in H2:H11, enter this in D2 to calculate their mean:

=AVERAGE(H2:H11)

With the same sample values in H2:H11, enter this in D3 to calculate the sample standard deviation:

=STDEV.S(H2:H11)

STDEV.S is appropriate when your data are a sample from a larger group. If the values include the entire population you want to describe, use STDEV.P instead. For this example, the sum of squared deviations is 36.9; dividing by the sample degrees of freedom, 9, gives a variance of 4.1, whose square root is approximately 2.0248. That is why STDEV.S returns about 2.02. The sample and population functions can give different results, so choose based on what the data represent rather than which result looks preferable.

Check that the cells contain numbers, not numbers stored as text, and decide how to handle blanks or invalid entries before calculating. Excel ignores blank cells in a referenced range, but an accidental text entry can mean a value is not included as expected. Standard deviation is also sensitive to outliers: one unusually large observation can substantially increase the result.

Build a normal curve in Excel

To create a smooth curve, calculate a sequence of horizontal-axis values and the corresponding normal-distribution density. In the example, keep the observations in H2:H11, the mean in D2, and the sample standard deviation in D3. In D4, set the lower end of the graph to three standard deviations below the smallest observation:

=MIN($H$2:$H$11)-3*$D$3

With the same layout, enter this in D5 to set the upper end three standard deviations above the largest observation:

=MAX($H$2:$H$11)+3*$D$3

In A2, enter the lower endpoint. In A3, with D4 and D5 holding the endpoints and A2 holding the first x-value, enter:

=A2+($D$5-$D$4)/100

Fill A3 down through A102 to create 101 evenly spaced x-values. Select A2:A102 and B2:B102, then choose Insert > Insert Scatter (X, Y) or Bubble Chart > Scatter with Smooth Lines. A scatter chart is important here because the x-values are numerical and unevenly spaced labels would distort the horizontal scale.

With the mean in D2, the standard deviation in D3, and an x-value in A2, enter this in B2:

=NORM.DIST(A2,$D$2,$D$3,FALSE)

Fill B2 down through B102, then use the two columns as the chart series. NORM.DIST with FALSE returns the probability density at each x-value, not the probability of one exact measurement. The curve is scaled so its total area is 1. In this example it will be centered at 11.1; its shape and spread reflect the standard deviation of about 2.02. The values D4 and D5 give a display range, not a claim that observations outside it are impossible.

Create a histogram of observed values

For the actual distribution rather than a theoretical curve, select the numeric observations and choose Insert > Insert Statistic Chart > Histogram in Excel for Microsoft 365 on Windows. Excel groups the measurements into bins. To change the grouping, right-click the horizontal axis, choose Format Axis, and adjust the bin width or number of bins under Axis Options. The bin choice affects how clearly clusters, gaps, and skew appear, so try a few sensible settings rather than treating one display as definitive.

A histogram is especially useful before adding a normal curve because it reveals whether the bell-shaped assumption looks plausible. If the histogram has a long tail, two peaks, or a few isolated values, a normal curve may be a poor summary. Excel versions and platforms can present chart commands somewhat differently; on Mac, look for Histogram under the Insert chart options, then use the axis-formatting pane to adjust bins. Excel for the web may offer fewer chart-formatting controls than the desktop app.

A density curve and a frequency histogram do not automatically share the same vertical scale. Density has total area 1, while histogram bars usually show counts. To compare them fairly, use a probability-density histogram or scale the curve’s density by the sample size and bin width; otherwise, a curve can appear too low simply because its units differ.

Add standard deviation error bars

For a chart of group averages, first calculate one mean for each group and a standard deviation for the observations in that group. Select the chart, then use Chart Design > Add Chart Element > Error Bars > More Error Bars Options. Excel’s built-in Standard Deviation option applies a calculated spread to the plotted series. It is a convenient visual setting, but it may not represent the separate within-group standard deviation you calculated for each category.

When each group has its own standard deviation, choose Custom in the error-bar options and specify the cells containing positive and negative error amounts. For symmetric standard-deviation bars, use the same standard-deviation range for both. The chart should identify what the bars represent in its title, legend, or note. Standard deviation bars show spread among observations; standard error bars show uncertainty in an estimated mean. They are not interchangeable, and neither one alone establishes whether two groups differ meaningfully.

Read the graph without overclaiming

On a normal curve, the mean marks the center and the standard deviation controls the width. A taller, narrower curve generally indicates less spread; a shorter, wider curve indicates more spread, assuming the curves represent distributions with the same total area. To mark the mean and one-standard-deviation limits, calculate mean minus standard deviation and mean plus standard deviation, then add vertical reference lines or label those x-axis positions.

Remember that a standard deviation is measured in the same units as the original data. If measurements are in centimeters, the standard deviation is in centimeters too. It is not the same as the standard error, which estimates how much a sample mean would vary across repeated samples. Also, a standard deviation is not a universal measure of how far every value lies from the mean: individual observations can be much farther away, particularly in skewed data.

Common problems and practical fixes

  • A flat or oddly shaped curve: Check that the x-values are numeric and increase evenly, and that the mean and standard deviation cells contain the intended results.
  • A curve that does not align with histogram bars: Check the vertical units. Density and frequency counts need compatible scaling before they can be compared.
  • A misleading bell shape: A normal curve is generated from two summary statistics, so it can look smooth even when the observations are not normally distributed. Review the histogram and the raw values.
  • Unexpectedly large spread: Look for data-entry errors and genuine outliers. Do not remove valid observations only to make the chart look more regular.
  • Different results from another calculation: Confirm whether that calculation uses a sample or population standard deviation and whether it includes the same observations.

Keep the source data, calculation cells, and chart labels clear. A reader should be able to tell whether the display shows a normal model, observed counts, or error bars, and whether the standard deviation is based on a sample or a full population.

Frequently Asked Questions

Does Excel have a standard deviation chart button?
No. Excel provides statistical charts such as histograms and chart error bars, but you choose the display and calculate the statistic that fits your question. A normal curve can be made with a scatter chart and NORM.DIST values.

Should I use STDEV.S or STDEV.P?
Use STDEV.S when your observations are a sample from a larger population. Use STDEV.P when the data contain the complete population you intend to describe.

How do I show one standard deviation above and below the mean?
Calculate mean minus standard deviation and mean plus standard deviation, then label those two x-axis values or add reference lines. On a normal curve, those bounds contain about 68% of values only when the distribution is approximately normal.

Is a histogram the same as a bell curve?
No. A histogram groups and counts observed values. A bell curve is a smooth distribution model, often calculated from a mean and standard deviation. You can display both, but their vertical scales must be made compatible.

Can I use a standard deviation graph for a small dataset?
You can calculate and plot the values, but a small sample may not reveal the population’s true shape. Show the observations or histogram as well, and avoid presenting a fitted normal curve as proof that the data are normally distributed.

Related

Try it: the Percentage Formula Generator: percent of, percent change, increase and decrease, with the formula.

More in Charts & pivot tables

Sources: Function behaviour and menu paths: Microsoft Support documentation for Excel. Method & sources

Spotted an error in this guide? Let us know.