diff --git a/src/uu/shuf/BENCHMARKING.md b/src/uu/shuf/BENCHMARKING.md
new file mode 100644
index 000000000..7607f04b4
--- /dev/null
+++ b/src/uu/shuf/BENCHMARKING.md
@@ -0,0 +1,28 @@
+# Benchmarking shuf
+
+`shuf` is a simple utility, but there are at least two important cases
+benchmark: with and without repetition.
+
+When benchmarking changes, make sure to always build with the `--release` flag.
+You can compare with another branch by compiling on that branch and than
+renaming the executable from `shuf` to `shuf.old`.
+
+## Without repetition
+
+By default, `shuf` samples without repetition. To benchmark only the
+randomization and not IO, we can pass the `-i` flag with a range of numbers to
+randomly sample from. An example of a command that works well for testing:
+
+```shell
+hyperfine --warmup 10 "target/release/shuf -i 0-10000000"
+```
+
+## With repetition
+
+When repetition is allowed, `shuf` works very differently under the hood, so it
+should be benchmarked separately. In this case we have to pass the `-n` flag or
+the command will run forever. An example of a hyperfine command is
+
+```shell
+hyperfine --warmup 10 "target/release/shuf -r -n 10000000 -i 0-1000"
+```