Excel Trendline Types and When to Use Each

Excel has six trendline types: linear, exponential, logarithmic, polynomial, power, and moving average. Choose one based on the shape and behavior of your data: a trendline can summarize a pattern, but the wrong type—or an unjustified forecast—can make the chart misleading.

What an Excel trendline shows

A trendline is a line or curve added to a chart to represent the general direction of a data series. Excel calculates it from the plotted values, rather than simply connecting every point as a regular chart line does. A trendline can make an overall relationship easier to see when individual observations fluctuate.

Trendlines are available for many two-dimensional charts, including scatter charts and common line, column, and bar charts. They are not available for every chart type: for example, Excel does not support them on pie, doughnut, radar, or three-dimensional charts. The precise options also depend on the selected chart and data.

Excel’s six choices are mathematical models, not six visual styles. The first five fit a particular kind of equation to the data. A moving average works differently: it smooths nearby observations to reduce short-term variation. None proves that one variable causes another, and a good-looking curve is not by itself evidence that the model is appropriate.

The six trendline types

Linear: A straight line that represents a constant rate of increase or decrease. Use it when the points follow a roughly straight pattern. For example, values of 10, 14, 18, and 22 at x-values 1, 2, 3, and 4 fit the line y = 4x + 6. The line rises by four units for each one-unit increase in x.
Exponential: A curve suited to a relationship that changes by a roughly constant percentage or ratio, rather than by a constant amount. It is often considered for growth or decay that accelerates or slows proportionally. Excel’s exponential fit requires positive y-values; it cannot fit a series containing zero or negative values.
Logarithmic: A curve that can rise or fall quickly at first and then level off, or show the reverse pattern. It may suit data where changes are large at low x-values and smaller at higher ones. The x-values must be positive for the logarithmic model.
Polynomial: A curved fit that can bend and turn. In Excel, you can set its order from 2 through 6. A second-order polynomial can form one broad bend; higher orders allow more turns and can follow more complicated patterns. That flexibility also makes it easier to overfit a small or noisy dataset.
Power: A curve used when the relationship follows a power pattern, such as a quantity changing with a power of another quantity. It can model some scaling relationships. Excel requires positive x- and y-values for this fit.
Moving average: A smoothed line based on averages of successive groups of observations. It helps reveal a broad direction when the data are bumpy, but it does not represent the same kind of fitted equation as the other five types. The period controls how many points contribute to each average: a longer period smooths more and reacts less quickly to changes.

How to add a trendline in Excel

In Excel for Microsoft 365 on Windows, create a chart and select the data series that should receive the trendline. Then use Chart Design > Add Chart Element > Trendline and choose a type, or select More Trendline Options to open the formatting pane and configure it. You can also select the chart, use the Chart Elements button beside it when available, and turn on Trendline.

In the formatting pane, choose the trendline type and adjust options that apply to it. For example, a polynomial trendline has an order setting, while a moving average has a period setting. To show the fitted equation or the R-squared value on the chart, select the corresponding display checkboxes in the pane. Excel may format the equation as a chart label, which you can reposition by dragging.

The chart interface differs somewhat on Mac and in Excel for the web, but the general workflow is to select the chart or data series and find the trendline option under the chart design or chart formatting controls. If a command is not visible, verify that the chart type supports trendlines and that you selected a data series rather than only the chart area. Ribbon names and available editing controls can vary by platform and update.

How to choose the right type

Start with the data pattern and the meaning of the measurements, not with the curve that looks most impressive. A scatter chart is especially useful when both variables are numerical and the x-values represent measured quantities, such as elapsed time or temperature. It plots the actual x-values. In a standard line chart, the horizontal axis is often a sequence of categories; equal spacing between categories does not necessarily represent equal intervals in time or measurement.

Try linear when equal changes in x correspond approximately to equal changes in y.
Consider exponential when the data change by a similar proportion over equal x-intervals, and all y-values are positive.
Consider logarithmic when the response changes sharply at first and then levels off, with positive x-values.
Use polynomial when the pattern has a defensible bend or turn that a straight line cannot represent. Begin with a low order.
Consider power when the variables are positive and the relationship is plausibly a scaling pattern.
Use a moving average to make short-term fluctuations less prominent, especially in an ordered series. Choose a period that makes sense for the observation frequency.

For instance, if monthly sales generally rise by about the same number of units each month, a linear trendline may be a reasonable summary. If a population or balance increases by a fairly steady percentage, an exponential model may be more plausible. If a series has regular seasonal peaks, however, none of these simple trendlines captures the seasonal cycle by itself. A trendline should match the question and the data-generating context, not merely produce a high R-squared value.

Equations, R-squared, and forecasts

When you enable Display Equation on chart, Excel shows the fitted model in an abbreviated form. The coefficients describe the fitted relationship in the units and scale of the plotted data. The displayed equation can be rounded, so it may not reproduce every plotted result exactly if you copy its visible coefficients into a calculation.

Display R-squared value on chart adds a measure of how closely the model fits the data under the model’s calculation. Values nearer 1 generally indicate a closer fit to the plotted observations, but a high value does not establish causation, validate assumptions, or guarantee accurate predictions. Adding polynomial terms can raise the fit statistic while making the model less reliable beyond the observed range.

For supported trendline types, the formatting pane also provides forward and backward forecast settings. These extend the fitted line beyond the available x-values; they do not add actual observations or account automatically for events that may change the pattern. Treat a forecast as an extrapolation, and be particularly cautious when the forecast goes much farther than the data or when the series has seasonality, structural breaks, or a small sample.

Important chart and data limitations

Trendlines depend on the values Excel interprets as the series’ x- and y-coordinates. In an XY scatter chart, the x-values are numeric coordinates. In a line chart, the horizontal axis typically represents categories, and Excel may fit the series using their order rather than the real spacing between dates or numbers. If your observations are irregularly spaced, a scatter chart with actual numeric x-values is usually the clearer choice for analyzing their relationship.

Some models have mathematical restrictions. Exponential trendlines need positive y-values; logarithmic trendlines need positive x-values; and power trendlines need positive x- and y-values. Missing, zero, or negative observations can therefore prevent a desired fit or change which points can be used. Check the source data and chart setup instead of assuming Excel has fitted every intended observation.

A moving average is also easy to misread. It smooths the series according to the order of its points, so the period should be interpreted in observations, not automatically in days or months. If data are weekly, a period of four refers to four plotted points; it does not necessarily mean a calendar month. Smoothing can hide a sudden shift as well as noise, so retain the original series for comparison.

Common mistakes and ways to avoid them

Choosing the highest-order curve: A sixth-order polynomial can bend to follow individual observations. Prefer the simplest type that describes a credible pattern, and compare the fitted line with the raw points.
Reading correlation as cause: A trendline describes an association in the plotted data. Other variables, coincidence, or the way data were collected may explain the pattern.
Forecasting too far: A model that summarizes past observations may fail when conditions change. Show the observed data clearly and label an extension as a forecast, not as known future values.
Using category spacing as measured time: A line chart can space categories evenly even when the dates are not evenly spaced. Use a scatter chart when actual numeric intervals matter.
Confusing smoothing with a fitted relationship: A moving average suppresses short-term variation; it is not interchangeable with a linear or curved regression trendline.
Relying on a rounded equation: The equation printed on the chart may have fewer digits than the underlying fit. Avoid treating its displayed rounding as exact precision.

Trendline versus a chart line or forecast

A regular line connecting data points shows the observed sequence, including its rises and falls. A trendline summarizes a pattern across that sequence. A moving average is a third idea: it smooths the sequence over a chosen window. Keeping these roles distinct helps readers understand whether a chart displays actual values, a fitted model, or a smoothed view.

A chart trendline is also not the only way to forecast in Excel. For a detailed forecast, you may need a method that reflects the data structure, such as seasonality, or a worksheet analysis designed for that purpose. A simple trendline is useful for a visual summary, but it does not automatically perform a full time-series analysis or supply uncertainty bounds for its extrapolation.

Frequently Asked Questions

How many trendline types does Excel have?
Excel provides six: linear, exponential, logarithmic, polynomial, power, and moving average. The model you can apply depends on the chart, the data, and the mathematical requirements of the selected type.

Which Excel trendline is best?
There is no universally best type. Choose one that fits the visible pattern and makes sense for the variables. A linear model is a sensible starting point for a roughly straight relationship, while a moving average is useful for smoothing an ordered series.

Why can’t I add a trendline in Excel?
The chart type may not support trendlines, or you may have selected the chart background instead of a data series. Select a supported two-dimensional chart and its series, then look under the chart design controls. Some mathematical trendlines also require positive values.

What does a moving average period mean in Excel?
It is the number of successive plotted points used in the moving-average calculation. A larger period usually creates a smoother line but can make it slower to show a genuine change. Interpret the period relative to the frequency of your observations.

Can I use an Excel trendline to predict future values?
You can extend supported trendlines with the forecast settings, but that extension assumes the fitted pattern continues. It is an extrapolation, not a guarantee. Consider the amount and quality of historical data, and whether seasonality or changing conditions make a simple trend unsuitable.

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.