Excel Trendline Only Some Data Points? Fix It

In Excel, a trendline is normally calculated for an entire chart series, not for individually selected points. To fit only some data points, create a separate chart series containing just those points, then add the trendline to that series; keep the full series on the chart if you still want to show all the data.

How Excel trendlines work

A trendline is a statistical fit for the data in a chart series. Excel uses the series’ values to calculate a curve or line, then draws that fit on the chart. Selecting one marker does not normally tell Excel to use only that marker, or a few selected markers, for the calculation. The trendline belongs to a series, so its input is the set of values in that series.

This distinction matters when a data set contains different phases. Imagine monthly sales for two years, where the first six months reflect an unusual launch and you want a trendline for the remaining months. If all 24 months are in one series, a trendline added to it uses the whole series. It will not automatically ignore the launch period just because you clicked a later marker.

A trendline also is not the same as a line connecting observations. A connecting line joins each point in order; a trendline summarizes a pattern and may not pass through any particular point. Excel can display both, which is useful when the original values are noisy but you want to show the overall direction.

Make a separate series for the points you want

The most dependable method is to give Excel a distinct series containing only the observations intended for the fit. For example, if dates are in A2:A25 and measurements are in B2:B25, and you want the fit for the last eight observations, copy those eight dates and measurements to a helper range in two unused columns. Include headers so you can identify the helper series later. Do not include the earlier observations in that helper range.

If the chart should still display every observation, add the helper range as a second series on the existing chart. The original series remains responsible for showing the complete data; the helper series supplies only the points used for the trendline. If you only need to show the selected observations, make the chart from the helper range instead.

Copy the X values and corresponding Y values for the chosen observations into adjacent helper columns. Keep each X value beside its correct Y value.
Click the existing chart. On the ribbon, select Chart Design > Select Data > Add.
In the Edit Series dialog, set the series name to the helper header, then select the helper Y values for Series values. For an XY (Scatter) chart, also set Series X values to the matching helper X values. Do not include the header cells in the value ranges.
Click OK to add the series. Select the helper series in the chart, then choose Chart Design > Add Chart Element > Trendline > More Trendline Options. Choose the trendline type you need.

For a line chart, Excel uses the category labels on the horizontal axis rather than a separate numeric X range in the same way as an XY scatter chart. If the horizontal values represent measured quantities or irregularly spaced dates, an XY (Scatter) chart is often a better choice because it uses numeric X values and their spacing. Choose Insert > Insert Scatter (X, Y) or Bubble Chart to create a scatter chart when appropriate.

Keep all points visible without duplicating markers

When you add a subset series to a chart that already contains the full series, the selected points appear twice: once in the full series and once in the helper series. That is usually harmless, but overlapping markers can look darker or cluttered. You can format the helper series so it contributes to the trendline without adding a second visible set of markers.

Select the helper series, open Format Data Series, and in the formatting pane set its marker options to None. For a scatter chart, set the helper series line to No line as well if you do not want it to draw connecting segments. The trendline is a separate chart element, so formatting the helper series does not require you to hide the trendline. If your goal is to emphasize the selected observations, leave the helper markers on and format them with a contrasting color instead.

Check the legend after adding the series. Excel may give it a generic name such as Series 2 if you did not supply a clear header. Rename the helper column or edit the series name through Chart Design > Select Data so readers can tell which points the trendline represents.

Choose a chart and trendline type that fit the data

For a straight-line relationship, choose a linear trendline. Other built-in choices include exponential, logarithmic, polynomial, power, and moving average. The right option depends on what the data means, not simply which curve looks closest to the points. For example, an exponential fit assumes a particular kind of proportional growth and cannot be used when the relevant values are zero or negative. A polynomial can bend to follow changes, but a high order can make random fluctuations look like a meaningful pattern.

Moving average trendlines smooth short-term variation rather than fitting a single straight relationship. They are useful for displaying a rolling direction, but the first plotted values depend on the selected period and may not appear in the same way as a standard regression trendline. If the chart does not offer the option you expect, check the chart type: trendlines are not available for every chart. Excel does not support them for chart types such as 3-D, stacked, radar, pie, or doughnut charts.

For a scatter chart with a numeric horizontal axis, a trendline is fitted using the X and Y values. In a line chart, the horizontal axis is generally categorical, so the spacing and interpretation can differ. If the dates are irregular—for example, measurements taken after 1, 2, and 20 days—a line chart can space categories evenly even though the elapsed time is not even. A scatter chart makes that time gap visible and is usually more appropriate for analyzing a numeric relationship.

Limit where the trendline appears

There are two different requests that sound like “trendline only some data points.” One is to calculate the fit from a subset, which requires the separate-series method. The other is to calculate a fit from the whole series but show only part of the drawn line. These are not equivalent: shortening the displayed line does not change the observations used in the calculation.

For a standard trendline, Excel draws the fit over the range represented by the series. In Format Trendline, the forecast settings can extend the line forward or backward beyond the data range; set those extensions to zero if you do not want a forecast. If you need a line that starts or stops at a particular internal point while still fitting a larger data set, Excel’s standard trendline controls may not provide that exact combination. A separately calculated series or a separate chart element may be needed, depending on the intended presentation.

Be explicit in the chart title, legend, or a nearby note about the interval used. A trendline based on only a selected period can suggest a different story from one based on all observations. Naming the range—for example, “Trend, post-launch months”—helps prevent readers from assuming the line summarizes every point.

Common reasons the result looks wrong

The trendline still includes every point. It may have been added to the original series rather than the helper series. Click the trendline and check which series it belongs to, or remove it and add one after selecting the helper series.
The selected points do not line up with their X values. Make sure each helper X value is copied from the same row as its Y value. Sorting or selecting the two columns separately can pair observations incorrectly.
The trendline extends beyond the intended interval. Check the forecast forward and backward settings in Format Trendline. Also confirm that the helper range contains only the intended observations.
Some points are absent or the fit is unexpected. Look for text stored in numeric cells, blank cells, errors, or an unintended chart range. Check the series ranges in Chart Design > Select Data.
The points are evenly spaced when they should not be. A line chart may treat dates or labels as categories. Consider changing to an XY (Scatter) chart when horizontal distances have numerical meaning.
The curve looks convincing but does not make sense. A more complex trendline can fit a small sample closely without describing a reliable pattern. Review the data and the assumptions of the selected trendline rather than choosing a type only for appearance.

Filtering or hiding rows is not as clear a way to define the trendline’s input. Chart settings can affect whether hidden data is displayed, and a filtered chart can be harder to maintain or explain. A helper series makes the exact included observations visible in the worksheet and easier to verify.

Check the fit and explain its limits

After adding the trendline, verify the result against the selected data. For a linear fit, does its direction match the observations? Are there outliers, a change in measurement method, or a break between two operating periods? A trendline is a compact summary, not evidence that one variable causes another or that the same pattern will continue.

Excel can display the trendline equation and the R-squared value through the options in the Format Trendline pane. These can help describe a fit, but they need context. R-squared is not a general score that proves a model is appropriate, and a high value does not guarantee a good forecast. The displayed equation may also be rounded, so using its printed coefficients for precise calculations can produce slightly different results from the underlying fit.

For a small number of points, consider whether a trendline is informative at all. Two points always determine a straight line, but that does not establish a stable trend. With a short or noisy subset, show the observations clearly and avoid presenting a fitted line as more certain than the data supports.

Frequently Asked Questions

Can I add a trendline to just one selected point in Excel?
Not as a meaningful trendline fit. A trendline applies to a chart series, and one point does not provide a useful basis for estimating a trend. Put the intended group of points in its own series and add the trendline to that series.

Can I select several markers with Ctrl and apply a trendline only to them?
Excel does not normally use a multi-selection of individual markers to define a trendline. Create a helper series containing those observations instead. This makes the subset explicit and keeps the trendline attached to the correct data.

Can I exclude outliers without deleting them?
Yes. Keep the original series to show all observations, then create a second series that omits the points you do not want included in the fit. Add the trendline to that second series and label the choice clearly. Excluding an outlier should have a defensible reason, not just make the line look better.

Why does my trendline change when I edit the chart data?
The trendline recalculates when the values or series range it belongs to change. If new rows become part of the helper range, they can affect the fit; if the helper range is fixed, new observations may be left out. Review the series ranges in Select Data whenever the source data changes.

Does this work in Excel for Mac or Excel for the web?
The separate-series approach is also applicable in Excel for Mac, though ribbon labels and dialog layouts can vary by release. Excel for the web supports chart editing, but some detailed formatting and trendline options may be more limited or arranged differently than in Microsoft 365 for Windows. If an option is missing, open the workbook in the desktop app.

How can I make a trendline start later but still use all the data?
That is a display-range request, not a subset-fit request. The standard trendline forecast controls extend the line beyond the series range, but they do not generally let you set an arbitrary internal start point for a fit calculated from all observations. Use a separate series or another chart element if you need that precise visual result.

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.