Written and maintained by CASRAI Editorial Board
Last updated
abline() is the base R graphics function that adds one or more straight lines to a plot that already exists: a line from an intercept and slope, a horizontal line at a y-value, a vertical line at an x-value, or the line from a fitted regression object. This guide is for researchers who need to mark a threshold, a mean or a fitted trend on a figure and get the call right the first time. It also covers the ggplot2 equivalents, because geom_abline(), geom_hline() and geom_vline() take different argument names.
What abline in R is used for in research
In practice abline() is used to overlay a reference on a figure that has already been drawn:
- a fitted least-squares line on a scatterplot (
abline(fit)), see the linear regression scatter plot guide; - a horizontal line at a cut-off, mean or limit (
h =); - a vertical line at a time point, dose or threshold (
v =), which is the usual answer to a “vertical line plot” question in base R; - a line of identity or any other line with a known intercept and slope (
a =,b =).
abline() draws on the current plot, so a plot must be created first, for example with plot(). It does not create a figure by itself. If you want to draw a line from your own data points rather than from a formula, see plotting a line in R.
Install or access
abline() belongs to the graphics package, which is part of base R, so nothing needs to be installed beyond R itself (available from the R Project). The ggplot2 equivalents need the ggplot2 package, installed with install.packages("ggplot2") and loaded with library(ggplot2).
abline() arguments
The documented usage in the R reference manual (graphics package) is:
abline(a = NULL, b = NULL, h = NULL, v = NULL, reg = NULL, coef = NULL, untf = FALSE, ...)| Argument | What it does | ggplot2 equivalent |
|---|---|---|
a, b |
Intercept and slope, each a single value | geom_abline(intercept =, slope =) |
h |
y-value(s) at which to draw horizontal line(s) | geom_hline(yintercept =) |
v |
x-value(s) at which to draw vertical line(s) | geom_vline(xintercept =) |
coef |
A vector of length two: intercept, then slope | geom_abline() with the two values |
reg |
An object with a coef method, such as an lm fit |
geom_abline() with the values from coef(fit) |
untf |
Only relevant when an axis is log-transformed (see below) | Not applicable |
col, lty, lwd, xpd, lend, ljoin, lmitre |
Graphical parameters passed through ... |
colour, linetype, linewidth, alpha |
Step-by-step: add a line with abline()
1. Draw the plot first
plot(mtcars$wt, mtcars$mpg)2. Add a line from an intercept and slope
abline(a = 1, b = 2, col = 2)This is the example form given in the reference manual: intercept 1, slope 2, drawn in colour 2.
3. Add a regression line
fit <- lm(mpg ~ wt, data = mtcars)abline(fit)Passing a fitted model is the same as abline(coef = coef(fit)), as the manual’s example notes for an lm fit. For an intercept-only model the manual draws the line with abline(a = coef(model), b = 0, col = "blue").
4. Add horizontal and vertical lines
abline(h = 20)abline(v = 3.5, lty = 2)Both h and v accept vectors, so abline(h = c(15, 20, 25)) draws three lines. col, lty and lwd can also be vectors for h and v, and they are recycled across the lines. Line width and type are explained in lwd in R and ggplot line types.
The same lines in ggplot2
p <- ggplot(mtcars, aes(wt, mpg)) + geom_point()p + geom_vline(xintercept = 5)p + geom_hline(yintercept = 20)p + geom_abline(intercept = 37, slope = -5)These are the examples in the ggplot2 reference. The three geoms do not inherit the plot’s x and y aesthetics and do not affect the scales. They are drawn using the same aesthetics as geom_line(): alpha, colour, linetype and linewidth.
Output and how to read it
abline() adds a line to the open graphics device; it does not return a new plot. Read the line against the axes of the existing plot: an h line crosses the y-axis at the value you gave, a v line crosses the x-axis at its value, and an a/b line rises by b y-units for each 1 x-unit. A fitted line from abline(fit) is a visual summary of the model, not a test of it: check the residuals and model assumptions separately, and see the scatterplots in R guide for the wider plotting workflow.
Log-transformed axes and untf
With untf = TRUE and a log-transformed axis, the line is drawn in the original coordinates, so it appears as a curve. Lines drawn with h and v always use the original coordinates.
Troubleshooting
abline() fails because no plot is open
abline() adds to the current plot, so with no plot open there is nothing to draw on. Call plot() (or another high-level plot function) first.
The ggplot2 line does not show
Lines outside the data range are invisible, and geom_abline() called with no arguments can look empty. Check that the intercept or slope values put the line inside the plotted range.
The line repeats in every facet
Arguments given directly to the geom draw the same line in every facet. To draw a different line per facet, put the values in a data frame and map them with aes().
abline(reg = fit) draws a line through the origin
This happens when the model’s coef method returns a single value (a slope-only fit). Fit the model with an intercept, or pass both values through coef, to get a line that does not pass through the origin.
Free alternatives and related functions
| Task | Base R | ggplot2 |
|---|---|---|
| Line from intercept and slope | abline(a, b) |
geom_abline(intercept, slope) |
| Horizontal reference line | abline(h = ) |
geom_hline(yintercept = ) |
| Vertical reference line | abline(v = ) |
geom_vline(xintercept = ) |
| Line from a fitted model | abline(fit) |
geom_abline() with values from coef(fit) |
FAQ
What does abline do in R?
abline() adds one or more straight lines to the current plot: from an intercept and slope, at given y-values (h), at given x-values (v), or from a regression object (reg or coef).
How do I add a vertical line to a plot in R?
Draw the plot, then call abline(v = x) with the x-value, for example abline(v = 3.5). In ggplot2 use geom_vline(xintercept = 3.5).
How do I add a regression line with abline in R?
Fit the model with lm() and pass it to abline(), for example abline(lm(mpg ~ wt, data = mtcars)). It is equivalent to abline(coef = coef(fit)).
Can I use abline with ggplot2?
Not the function itself. ggplot2 provides geom_abline(), geom_hline() and geom_vline(), which are added to a plot with +.
Sources (official documentation)
- R documentation, graphics package: abline: Add Straight Lines to a Plot (read 2026-11-08).
- ggplot2 reference: Reference lines: horizontal, vertical, and diagonal (read 2026-11-08).
