Add Error Bars to an Excel Chart: A Practical Guide

Charts & pivot tables · Updated November 19, 2026

To add error bars in Excel for Microsoft 365 on Windows, select the chart, then choose Chart Design > Add Chart Element > Error Bars and select an option such as Standard Error, Percentage, or Standard Deviation. For values you calculate yourself, choose More Error Bars Options, then use Custom > Specify Value in the Format Error Bars pane.

What error bars show

Error bars are lines attached to chart points that show a chosen amount of spread, uncertainty, or measurement error. They can extend above and below a point, to one side, or horizontally as well as vertically on a scatter chart. The bars do not explain what that amount means by themselves: the chart title, axis labels, legend, or caption should tell readers whether they represent standard deviation, a confidence interval, a measurement tolerance, or something else.

For example, suppose a column chart shows the average time to complete a task for five groups. A bar extending 2 seconds above and below one group’s average could represent a 2-second measurement tolerance, a standard error, or a confidence interval. Those are different quantities, even if the drawn bars happen to be the same length. Choose the calculation based on the question and how the data were collected rather than selecting an option just because it is available in the menu.

Add a built-in error bar option

Excel can add commonly used error bars without requiring a separate worksheet calculation. First select the chart, and then select the data series whose values should receive the bars. If the chart has several series, make sure you are adding bars to the intended series; you can format other series separately.

  1. Select the chart. If you click a particular column, marker, or line, Excel may select just that series; check the selection before changing it.
  2. On the ribbon, choose Chart Design > Add Chart Element > Error Bars.
  3. Choose Standard Error, Percentage, or Standard Deviation, or choose More Error Bars Options to open the Format Error Bars pane and adjust settings.
  4. Review the chart at a readable size. If the bars obscure the data or dominate the chart, check that the option matches the scale and purpose of the data.

You can also use the Chart Elements button, shown as a plus sign beside a selected chart, and turn on Error Bars. The available choices can depend on the chart type and the selected series. On Excel for Mac, select the chart and use the Chart Design tab’s chart-element controls; pane placement and some labels can differ slightly. Excel for the web has a more limited chart-editing interface in some cases, so open the workbook in the desktop app if you cannot find the required error-bar controls.

Choose the right error-bar calculation

The built-in choices are convenient, but they answer different questions. Excel applies a calculated amount to the plotted series; they are not a substitute for knowing what each observation represents.

  • Standard Error: Shows a standard-error amount based on the values in the plotted series. It is often used to describe uncertainty in an estimate, but it is not automatically a 95% confidence interval. A confidence interval usually requires an appropriate multiplier and assumptions about the data.
  • Standard Deviation: Shows the spread of values around the series average. It can help communicate variability among observations, but it does not directly show how precisely the average is estimated.
  • Percentage: Sets the error amount as a percentage of each plotted value. This is useful when a relative allowance is intended, such as a consistent percentage tolerance. It is not a statistical calculation of uncertainty.
  • Fixed Value: Uses the same error amount for each point in the selected series. This can represent a constant instrument resolution or a shared, known tolerance.
  • Custom: Lets you supply a different positive and negative amount for each point. Use it for calculated intervals, asymmetric uncertainty, or values determined outside the chart.

Standard deviation and standard error are especially easy to confuse. If repeated measurements vary widely but the average is estimated from many observations, the standard deviation describes that variation, while the standard error describes uncertainty in the estimated average. Neither one should be described as a confidence interval unless the interval has actually been calculated that way.

Set custom positive and negative values

Custom error bars are the most flexible option when you already have the uncertainty values in cells. Arrange the worksheet so the error amounts line up with the plotted points. For a five-point series, for instance, put five positive amounts in one range and five negative amounts in another. These cells should contain error distances, not the upper and lower endpoint values of the interval.

  1. Select the chart, then select the series to format.
  2. Choose Chart Design > Add Chart Element > Error Bars > More Error Bars Options. If the chart already has error bars, selecting them and opening the formatting pane may be quicker.
  3. In the Format Error Bars pane, find Error Amount, choose Custom, and select Specify Value.
  4. In the dialog, specify the cell range for Positive Error Value and the range for Negative Error Value. You can type a reference or select the cells on the worksheet.
  5. Confirm the dialog and check that the bars correspond to the intended points. Repeat for other series if needed.

For example, if the plotted series has five means and the upper and lower uncertainties differ, the positive range supplies the upward distances and the negative range supplies the downward distances. The number of values should match the number of plotted points. A common mistake is entering upper limits such as 14 when the plotted mean is 10 and the intended upward error is 4; Excel expects the distance, 4, not the endpoint, 14. If the uncertainty is symmetric, supply the same set of distances in both boxes.

Choose direction and format

In the Format Error Bars pane, settings such as direction and end style control how the bars look. Direction can be set to both sides, positive only, or negative only; the end style determines whether a small cap appears at the end. These settings affect display, not the underlying calculation. A one-sided bar can be appropriate when only an upper or lower bound is meaningful, but label that choice clearly so readers do not mistake it for a symmetric interval.

On an XY (scatter) chart, horizontal error bars can represent uncertainty in the X values, while vertical bars represent uncertainty in the Y values. If both are relevant, format each direction separately. In the Format Error Bars pane, use its selection control to choose the X or Y error bars before changing the settings. This distinction matters when, for example, both the measured concentration and the response have known measurement uncertainty. A regular column or line chart typically emphasizes vertical error bars because its horizontal positions are categories rather than measured numeric values.

Make the chart easier to interpret

Good error bars should be visible without making the data difficult to read. Use a line weight and color that contrast with the series but do not overpower it. If the chart has many closely spaced points, narrow bars or subtle caps may reduce clutter. Avoid relying on color alone to distinguish series, especially if the chart may be printed in grayscale.

Also explain what the bars mean. A short note such as “bars show standard deviation” or “bars show 95% confidence intervals” is more useful than a generic “error bars” label. If the bars represent custom uncertainty, state the source or method where space allows. For a chart in a report, the explanation can appear in a caption rather than inside the plot area.

Do not use overlap or non-overlap as a formal significance test. Whether two intervals overlap is not, by itself, a complete test of whether two estimates differ. The correct interpretation depends on the kind of interval, the study design, and the statistical method. Error bars are a visual aid; they do not replace an appropriate analysis.

Common problems and fixes

  • The bars are much too large or too small: Check whether you selected Percentage, Standard Deviation, or Standard Error unintentionally. For custom values, confirm that the cells contain distances and use the same units as the chart data.
  • Only one series has bars: Excel formats error bars by series. Select another series and add or configure its bars separately.
  • The custom values do not line up: Confirm that each error range has one value for every plotted point, in the same order as the series. If the chart filters or rearranges data, recheck the mapping.
  • Some bars are hard to see: They may be hidden behind markers, clipped by the axis limits, or too small at the current scale. Widen the plot area or adjust the axis range only when that change does not mislead readers.
  • The Error Bars command is missing: Confirm that a supported two-dimensional chart and a data series are selected. Error bars are commonly used with column, bar, line, and scatter charts; some chart types, including 3-D charts, do not offer the same support.

If you change the underlying data, revisit custom ranges and the visual scale. Cell-linked custom amounts can update when their source cells change, but manually supplied values or changed point order may no longer describe the intended uncertainty.

Frequently Asked Questions

How do I add error bars to just one series in Excel?
Select the chart and then the specific series before choosing Chart Design > Add Chart Element > Error Bars. Excel applies the bars to the selected series, so repeat the process for any other series that needs a different treatment.

Can I make different error bars for each data point?
Yes. Choose More Error Bars Options > Custom > Specify Value and provide cell ranges containing one error distance per point. You can use different ranges for positive and negative values to create asymmetric bars.

Should I use standard deviation or standard error?
Use standard deviation when you want to show variation among observations. Standard error concerns uncertainty in an estimated statistic, commonly a mean. Choose based on what the data and analysis support, and label the bars so readers know which quantity is shown.

Can Excel show horizontal error bars?
Yes, on suitable charts such as XY scatter charts. Select the X error bars in the Format Error Bars pane to represent uncertainty in horizontal values; use Y error bars for vertical values.

Why do error bars look different for negative values?
Bars are drawn from each plotted point, so their visual direction follows the chart’s axes and the selected positive or negative setting. Check the source values, axis scale, and custom positive and negative ranges if the result seems unexpected.

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.