diff --git a/docs/source/acknowledgements.rst b/docs/source/acknowledgements.rst
index 9a94fb73..48b240cb 100644
--- a/docs/source/acknowledgements.rst
+++ b/docs/source/acknowledgements.rst
@@ -20,8 +20,9 @@ based on the approach described in
 
 Checkpointing with tlm_adjoint is described in
 
-- James R. Maddison, 'On the implementation of checkpointing with high-level
-  algorithmic differentiation', https://arxiv.org/abs/2305.09568v1, 2023
+- James R. Maddison, 'Step-based checkpointing with high-level algorithmic
+  differentiation', Journal of Computational Science 82, 102405, 2024,
+  doi: 10.1016/j.jocs.2024.102405
 
 References
 ----------
diff --git a/tlm_adjoint/checkpoint_schedules/mixed.py b/tlm_adjoint/checkpoint_schedules/mixed.py
index 05aa687a..6ada77f2 100644
--- a/tlm_adjoint/checkpoint_schedules/mixed.py
+++ b/tlm_adjoint/checkpoint_schedules/mixed.py
@@ -148,11 +148,11 @@ class MixedCheckpointSchedule(CheckpointSchedule):
     required to restart the forward has the same size as the data required to
     advance the adjoint over a step.
 
-    An updated version of the algorithm described in
+    Described in
 
-        - James R. Maddison, 'On the implementation of checkpointing with
-          high-level algorithmic differentiation',
-          https://arxiv.org/abs/2305.09568v1, 2023
+        - James R. Maddison, 'Step-based checkpointing with high-level
+          algorithmic differentiation', Journal of Computational Science 82,
+          102405, 2024, doi: 10.1016/j.jocs.2024.102405
 
     Offline, one adjoint calculation permitted.
 
diff --git a/tlm_adjoint/checkpoint_schedules/schedule.py b/tlm_adjoint/checkpoint_schedules/schedule.py
index ccb0a2f8..cb70f396 100644
--- a/tlm_adjoint/checkpoint_schedules/schedule.py
+++ b/tlm_adjoint/checkpoint_schedules/schedule.py
@@ -264,9 +264,9 @@ def action_forward(cp_action):
     dependency data either for storage in checkpointing units or for immediate
     use by the adjoint. For details see
 
-        - James R. Maddison, 'On the implementation of checkpointing with
-          high-level algorithmic differentiation',
-          https://arxiv.org/abs/2305.09568v1, 2023
+        - James R. Maddison, 'Step-based checkpointing with high-level
+          algorithmic differentiation', Journal of Computational Science 82,
+          102405, 2024, doi: 10.1016/j.jocs.2024.102405
 
     In 'offline' schedules, where the number of steps in the forward
     calculation is initially known, this should be provided using the `max_n`