diff --git a/tests/bench_runtime_coreapis/Makefile b/tests/bench_runtime_coreapis/Makefile
new file mode 100644
index 0000000000..14f75fd6b7
--- /dev/null
+++ b/tests/bench_runtime_coreapis/Makefile
@@ -0,0 +1,12 @@
+include ../Makefile.tests_common
+
+# we use thread flags in this benchmark by default, disable on demand
+USEMODULE += core_thread_flags
+USEMODULE += benchmark
+
+TEST_ON_CI_WHITELIST += all
+
+include $(RIOTBASE)/Makefile.include
+
+test:
+	tests/01-run.py
diff --git a/tests/bench_runtime_coreapis/README.md b/tests/bench_runtime_coreapis/README.md
new file mode 100644
index 0000000000..3a345451b9
--- /dev/null
+++ b/tests/bench_runtime_coreapis/README.md
@@ -0,0 +1,7 @@
+# Measure Runtime of Selected Core API functions
+
+This benchmark application measures the runtime of selected core API functions.
+Its purpose is to provide a baseline to assess the impacts when doing changes to
+core code.
+
+This application is not complete, simply add additional runs if needed.
diff --git a/tests/bench_runtime_coreapis/main.c b/tests/bench_runtime_coreapis/main.c
new file mode 100644
index 0000000000..5692d5e4cd
--- /dev/null
+++ b/tests/bench_runtime_coreapis/main.c
@@ -0,0 +1,83 @@
+/*
+ * Copyright (C) 2018 Freie Universität Berlin
+ *
+ * This file is subject to the terms and conditions of the GNU Lesser
+ * General Public License v2.1. See the file LICENSE in the top level
+ * directory for more details.
+ */
+
+/**
+ * @ingroup     tests
+ * @{
+ *
+ * @file
+ * @brief       Measure runtime of select core API functions
+ *
+ * @author      Hauke Petersen <hauke.petersen@fu-berlin.de>
+ *
+ * @}
+ */
+
+#include <stdio.h>
+
+#include "mutex.h"
+#include "benchmark.h"
+#include "thread.h"
+#include "thread_flags.h"
+
+#ifndef BENCH_RUNS
+#define BENCH_RUNS          (1000UL * 1000UL)
+#endif
+
+static mutex_t _lock;
+static thread_t *t;
+static thread_flags_t _flag = 0x0001;
+static msg_t _msg;
+
+static void _mutex_lockunlock(void)
+{
+    mutex_lock(&_lock);
+    mutex_unlock(&_lock);
+}
+
+static void _flag_waitany(void)
+{
+    thread_flags_set(t, _flag);
+    thread_flags_wait_any(_flag);
+}
+
+static void _flag_waitall(void)
+{
+    thread_flags_set(t, _flag);
+    thread_flags_wait_all(_flag);
+}
+
+static void _flag_waitone(void)
+{
+    thread_flags_set(t, _flag);
+    thread_flags_wait_one(_flag);
+}
+
+int main(void)
+{
+    puts("Runtime of Selected Core API functions\n");
+
+    t = (thread_t *)sched_active_thread;
+
+    BENCHMARK_FUNC("nop loop", BENCH_RUNS, __asm__ volatile ("nop"));
+    puts("");
+    BENCHMARK_FUNC("mutex_init()", BENCH_RUNS, mutex_init(&_lock));
+    BENCHMARK_FUNC("mutex lock/unlock", BENCH_RUNS, _mutex_lockunlock());
+    puts("");
+    BENCHMARK_FUNC("thread_flags_set()", BENCH_RUNS, thread_flags_set(t, _flag));
+    BENCHMARK_FUNC("thread_flags_clear()", BENCH_RUNS, thread_flags_clear(_flag));
+    BENCHMARK_FUNC("thread flags set/wait any", BENCH_RUNS, _flag_waitany());
+    BENCHMARK_FUNC("thread flags set/wait all", BENCH_RUNS, _flag_waitall());
+    BENCHMARK_FUNC("thread flags set/wait one", BENCH_RUNS, _flag_waitone());
+    puts("");
+    BENCHMARK_FUNC("msg_try_receive()", BENCH_RUNS, msg_try_receive(&_msg));
+    BENCHMARK_FUNC("msg_avail()", BENCH_RUNS, msg_avail());
+
+    puts("\n[SUCCESS]");
+    return 0;
+}
diff --git a/tests/bench_runtime_coreapis/tests/01-run.py b/tests/bench_runtime_coreapis/tests/01-run.py
new file mode 100755
index 0000000000..d741442230
--- /dev/null
+++ b/tests/bench_runtime_coreapis/tests/01-run.py
@@ -0,0 +1,24 @@
+#!/usr/bin/env python3
+
+# Copyright (C) 2018 Freie Universität Berlin
+#
+# This file is subject to the terms and conditions of the GNU Lesser
+# General Public License v2.1. See the file LICENSE in the top level
+# directory for more details.
+
+import os
+import sys
+
+
+# The default timeout is not enough for this test on some of the slower boards
+TIMEOUT = 30
+
+
+def testfunc(child):
+    child.expect_exact('[SUCCESS]', timeout=TIMEOUT)
+
+
+if __name__ == "__main__":
+    sys.path.append(os.path.join(os.environ['RIOTTOOLS'], 'testrunner'))
+    from testrunner import run
+    sys.exit(run(testfunc))
diff --git a/tests/periph_gpio/tests/02-bench.py b/tests/periph_gpio/tests/02-bench.py
index 954089ea06..e1cc79c21f 100755
--- a/tests/periph_gpio/tests/02-bench.py
+++ b/tests/periph_gpio/tests/02-bench.py
@@ -16,12 +16,12 @@ def testfunc(child):
 
     for pin in range(0, 8):
         child.sendline("bench 0 {}".format(pin))
-        child.expect(r" *nop loop: +(\d+)us  --- +(\d+\.\d+)us per call")
-        child.expect(r" *gpio_set: +(\d+)us  --- +(\d+\.\d+)us per call")
-        child.expect(r" *gpio_clear: +(\d+)us  --- +(\d+\.\d+)us per call")
-        child.expect(r" *gpio_toggle: +(\d+)us  --- +(\d+\.\d+)us per call")
-        child.expect(r" *gpio_read: +(\d+)us  --- +(\d+\.\d+)us per call")
-        child.expect(r" *gpio_write: +(\d+)us  --- +(\d+\.\d+)us per call")
+        child.expect(r" *nop loop: +(\d+)us  --- +(\d+\.\d+)us per call  --- +(\d+) calls per sec")
+        child.expect(r" *gpio_set: +(\d+)us  --- +(\d+\.\d+)us per call  --- +(\d+) calls per sec")
+        child.expect(r" *gpio_clear: +(\d+)us  --- +(\d+\.\d+)us per call  --- +(\d+) calls per sec")
+        child.expect(r" *gpio_toggle: +(\d+)us  --- +(\d+\.\d+)us per call  --- +(\d+) calls per sec")
+        child.expect(r" *gpio_read: +(\d+)us  --- +(\d+\.\d+)us per call  --- +(\d+) calls per sec")
+        child.expect(r" *gpio_write: +(\d+)us  --- +(\d+\.\d+)us per call  --- +(\d+) calls per sec")
         child.expect_exact(" --- DONE ---")
         child.expect_exact(">")