diff --git a/src/roboticstoolbox/backends/PyPlot/PyPlot.py b/src/roboticstoolbox/backends/PyPlot/PyPlot.py
index bdde273c7..a866d67e4 100644
--- a/src/roboticstoolbox/backends/PyPlot/PyPlot.py
+++ b/src/roboticstoolbox/backends/PyPlot/PyPlot.py
@@ -283,6 +283,8 @@ def step(self, dt: float = 0.05) -> None:
 
         :param dt: time step in seconds, defaults to 50 (0.05 s)
         :type dt: int, optional
+        :raises ValueError: if ``dt`` is not greater than zero in a window, where the
+            plot pauses for ``dt``
 
         ``env.step(args)`` triggers an update of the 3D scene in the matplotlib
         window referenced by ``env``.
@@ -295,10 +297,20 @@ def step(self, dt: float = 0.05) -> None:
               updated in the internal state of the robot object.
             - The control type is defined by the robot object, and not all
               robot objects support all control types.
-            - Execution is blocked for the specified interval
+            - In a window the execution is blocked for the specified interval, which
+              must be greater than zero: Matplotlib's ``pause(0)`` does not return,
+              since a timeout of zero means to wait until the event loop is stopped.
+              In a notebook there is no pause and ``dt`` of zero is fine.
 
+        .. versionchanged:: 1.5.0
+            In a window, ``dt`` of zero or less raises ``ValueError``, it used to wait
+            for ever.
         """
-
+        if dt <= 0 and self.render_mode == "window":
+            raise ValueError(
+                f"dt is {dt}, it must be greater than zero: a pause of zero seconds "
+                "does not return"
+            )
 
         # update the robot's state
         for rpl in self.robots:
diff --git a/src/roboticstoolbox/backends/PyPlot/PyPlot2.py b/src/roboticstoolbox/backends/PyPlot/PyPlot2.py
index 376322d88..c17aa3b5f 100644
--- a/src/roboticstoolbox/backends/PyPlot/PyPlot2.py
+++ b/src/roboticstoolbox/backends/PyPlot/PyPlot2.py
@@ -191,8 +191,17 @@ def step(self, dt: float = 0.05) -> None:
         by the robot object, and not all robot objects support all control
         types.
 
+        :param dt: time step in seconds, which must be greater than zero in a window,
+            where the plot pauses for ``dt``; in a notebook there is no pause
+        :raises ValueError: if ``dt`` is not greater than zero in a window:
+            Matplotlib's ``pause(0)`` does not return, since a timeout of zero means
+            to wait until the event loop is stopped
         """
-
+        if dt <= 0 and self.render_mode == "window":
+            raise ValueError(
+                f"dt is {dt}, it must be greater than zero: a pause of zero seconds "
+                "does not return"
+            )
 
         self._step_robots(dt)
 
diff --git a/tests/test_step_dt.py b/tests/test_step_dt.py
new file mode 100644
index 000000000..a5608d6f3
--- /dev/null
+++ b/tests/test_step_dt.py
@@ -0,0 +1,82 @@
+"""
+A time step of zero used to hang: ``plt.pause(0)`` waits until the event loop is stopped.
+"""
+
+import matplotlib
+
+matplotlib.use("Agg")
+
+import matplotlib.pyplot as plt  # noqa: E402
+import pytest  # noqa: E402
+
+import roboticstoolbox as rtb  # noqa: E402
+from roboticstoolbox import ET2  # noqa: E402
+from roboticstoolbox.backends.PyPlot import PyPlot, PyPlot2  # noqa: E402
+
+pytestmark = pytest.mark.timeout(20)
+
+
+@pytest.fixture(autouse=True)
+def close_figures():
+    plt.close("all")
+    yield
+    plt.close("all")
+
+
+@pytest.fixture
+def env3d():
+    env = PyPlot()
+    env.launch()
+    env.add(rtb.models.DH.Puma560(), readonly=True)
+    return env
+
+
+@pytest.fixture
+def env2d():
+    env = PyPlot2()
+    env.launch()
+    env.add(rtb.Robot2(ET2.R() * ET2.tx(1) * ET2.R() * ET2.tx(1)), readonly=True)
+    return env
+
+
+@pytest.mark.parametrize("dt", [0, 0.0, -0.05])
+def test_3d_step_needs_a_positive_dt(env3d, dt):
+    with pytest.raises(ValueError, match="greater than zero"):
+        env3d.step(dt)
+
+
+@pytest.mark.parametrize("dt", [0, 0.0, -0.05])
+def test_2d_step_needs_a_positive_dt(env2d, dt):
+    with pytest.raises(ValueError, match="greater than zero"):
+        env2d.step(dt)
+
+
+def inline_env(cls, robot):
+    env = cls()
+    env.launch(render_mode="notebook-inline")
+    env.add(robot, readonly=True)
+    return env
+
+
+@pytest.mark.parametrize("dt", [0, 0.0])
+def test_in_a_notebook_there_is_no_pause_and_zero_is_fine(dt):
+    # the notebook animation draws a frame for each step and does not pause
+    inline_env(PyPlot, rtb.models.DH.Puma560()).step(dt)
+    inline_env(
+        PyPlot2, rtb.Robot2(ET2.R() * ET2.tx(1) * ET2.R() * ET2.tx(1))
+    ).step(dt)
+
+
+def test_a_small_dt_is_fine(env3d, env2d):
+    env3d.step(0.001)
+    env2d.step(0.001)
+
+
+def test_the_default_dt_is_fine(env3d):
+    env3d.step()
+
+
+def test_plot_with_a_zero_dt_raises_and_does_not_hang():
+    puma = rtb.models.DH.Puma560()
+    with pytest.raises(ValueError, match="greater than zero"):
+        puma.plot(puma.qr, backend="pyplot", dt=0, block=False)