diff --git a/tests/threads.rs b/tests/threads.rs
new file mode 100644
index 0000000..db6fb03
--- /dev/null
+++ b/tests/threads.rs
@@ -0,0 +1,54 @@
+use std::sync::mpsc;
+use std::thread;
+use std::time::Duration;
+
+use r2r::*;
+
+#[test]
+// Let's create and drop a lot of node and publishers for a while to see that we can cope.
+fn doesnt_crash() -> Result<(), ()> {
+    // a global shared context.
+    let ctx = Context::create()?;
+
+    for c in 0..10 {
+        let mut ths = Vec::new();
+        for i in 0..10 {
+            let ctx = ctx.clone();
+            ths.push(thread::spawn(move || {
+                let mut node = Node::create(ctx, &format!("testnode{}", i), "").unwrap();
+                let p = node
+                    .create_publisher::<std_msgs::msg::String>(&format!("/r2r{}", i))
+                    .unwrap();
+
+                let to_send = std_msgs::msg::String {
+                    data: format!("[node{}]: {}", i, c),
+                };
+
+                // move publisher to its own thread and publish as fast as we can
+                thread::spawn(move || {
+                    loop {
+                        let res = p.publish(&to_send);
+                        match res {
+                            Ok(_) => (),
+                            Err(_) => {
+                                // println!("publisher died, quitting thread.");
+                                break;
+                            }
+                        }
+                    }
+                });
+
+                // wait for 1 sec before dropping
+                std::thread::sleep(Duration::from_millis(1000));
+                // println!("all done {}-{}", c, i);
+            }));
+        }
+
+        for t in ths {
+            t.join().unwrap();
+        }
+        // println!("all threads done {}", c);
+    }
+
+    Ok(())
+}