diff --git a/Makefile b/Makefile
index d86741d90a4..14f28d517ae 100644
--- a/Makefile
+++ b/Makefile
@@ -17,11 +17,46 @@ else
 	@echo "benchmark tool already installed..."
 endif
 
+# Benchmark make targets extend scripts/benchmark_test.sh (see #16467).
+# Override defaults with ARGS, e.g.:
+#   make bench-put ARGS='--total=100000 --clients=100'
+#   make bench-range ARGS='--total=50000 --consistency=s'
+
 .PHONY: bench-put
 bench-put: build install-benchmark
 	@echo "Running benchmark: put $(ARGS)"
 	./scripts/benchmark_test.sh put $(ARGS)
 
+.PHONY: bench-range
+bench-range: build install-benchmark
+	@echo "Running benchmark: range $(ARGS)"
+	./scripts/benchmark_test.sh range foo $(ARGS)
+
+.PHONY: bench-txn-put
+bench-txn-put: build install-benchmark
+	@echo "Running benchmark: txn-put $(ARGS)"
+	./scripts/benchmark_test.sh txn-put $(ARGS)
+
+.PHONY: bench-stm
+bench-stm: build install-benchmark
+	@echo "Running benchmark: stm $(ARGS)"
+	./scripts/benchmark_test.sh stm $(ARGS)
+
+.PHONY: bench-lease-keepalive
+bench-lease-keepalive: build install-benchmark
+	@echo "Running benchmark: lease-keepalive $(ARGS)"
+	./scripts/benchmark_test.sh lease-keepalive $(ARGS)
+
+.PHONY: bench-watch
+bench-watch: build install-benchmark
+	@echo "Running benchmark: watch $(ARGS)"
+	./scripts/benchmark_test.sh watch $(ARGS)
+
+.PHONY: bench-watch-latency
+bench-watch-latency: build install-benchmark
+	@echo "Running benchmark: watch-latency $(ARGS)"
+	./scripts/benchmark_test.sh watch-latency $(ARGS)
+
 PLATFORMS=linux-amd64 linux-386 linux-arm linux-arm64 linux-ppc64le linux-s390x darwin-amd64 darwin-arm64 windows-amd64 windows-arm64
 
 .PHONY: build-all
diff --git a/tools/benchmark/README.md b/tools/benchmark/README.md
index c3dbe919ee8..22837a0c10a 100644
--- a/tools/benchmark/README.md
+++ b/tools/benchmark/README.md
@@ -22,6 +22,38 @@ Alternatively, instead of installing the tool, you can use it by simply running
   $ go run ./tools/benchmark
 ```
 
+## Running via make (local etcd)
+
+From the repository root, use the `bench-*` make targets. Each target builds etcd,
+installs the benchmark tool if needed, starts a temporary local etcd process, runs
+the named workload with `--report-perfdash`, then tears the server down
+(see `scripts/benchmark_test.sh` and
+[#16467](https://github.com/etcd-io/etcd/issues/16467)).
+
+```bash
+make bench-put
+make bench-range
+make bench-txn-put
+make bench-stm
+make bench-lease-keepalive
+make bench-watch
+make bench-watch-latency
+```
+
+Pass extra flags through `ARGS`:
+
+```bash
+make bench-put ARGS='--total=100000 --clients=100 --conns=10'
+make bench-range ARGS='--total=50000 --consistency=s --limit=100'
+```
+
+`bench-range` uses a default key of `foo`. For custom range bounds, call the
+script directly:
+
+```bash
+./scripts/benchmark_test.sh range mykey myend --total=10000
+```
+
 ## Usage
 
 The following command should output the usage per the latest development.