diff --git a/R/README.md b/R/README.md
index 31174c73526f..da9f042b4fde 100644
--- a/R/README.md
+++ b/R/README.md
@@ -17,10 +17,14 @@ export R_HOME=/home/username/R
 
 #### Build Spark
 
-Build Spark with [Maven](https://spark.apache.org/docs/latest/building-spark.html#buildmvn) and include the `-Psparkr` profile to build the R package. For example to use the default Hadoop versions you can run
+Build Spark with [Maven](https://spark.apache.org/docs/latest/building-spark.html#buildmvn) or [SBT](https://spark.apache.org/docs/latest/building-spark.html#building-with-sbt), and include the `-Psparkr` profile to build the R package. For example to use the default Hadoop versions you can run
 
 ```bash
+# Maven
 ./build/mvn -DskipTests -Psparkr package
+
+# SBT
+./build/sbt -Psparkr package
 ```
 
 #### Running sparkR
diff --git a/project/SparkBuild.scala b/project/SparkBuild.scala
index 54ac3c19fa83..b872668db765 100644
--- a/project/SparkBuild.scala
+++ b/project/SparkBuild.scala
@@ -414,6 +414,10 @@ object SparkBuild extends PomBuild {
 
   enable(YARN.settings)(yarn)
 
+  if (profiles.contains("sparkr")) {
+    enable(SparkR.settings)(core)
+  }
+
   /**
    * Adds the ability to run the spark shell directly from SBT without building an assembly
    * jar.
@@ -888,6 +892,25 @@ object PySparkAssembly {
 
 }
 
+object SparkR {
+  import scala.sys.process.Process
+
+  val buildRPackage = taskKey[Unit]("Build the R package")
+  lazy val settings = Seq(
+    buildRPackage := {
+      val command = baseDirectory.value / ".." / "R" / "install-dev.sh"
+      Process(command.toString).!!
+    },
+    (Compile / compile) := (Def.taskDyn {
+      val c = (Compile / compile).value
+      Def.task {
+        (Compile / buildRPackage).value
+        c
+      }
+    }).value
+  )
+}
+
 object Unidoc {
 
   import BuildCommons._