diff --git a/CHANGELOG.md b/CHANGELOG.md
index 300b8559..6c439bf5 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -8,6 +8,7 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 ## [unreleased](https://github.com/mllam/neural-lam/compare/v0.6.0...HEAD)
 
 ### Added
+- Add `--num_sanity_val_steps` CLI argument to control sanity validation steps before training (#694)
 
 - Add `PropagationNet` GNN layer that incentivises directional message
   propagation from sender to receiver nodes, and expose it alongside
diff --git a/neural_lam/train_model.py b/neural_lam/train_model.py
index f98065c4..d4687c34 100644
--- a/neural_lam/train_model.py
+++ b/neural_lam/train_model.py
@@ -240,6 +240,13 @@ def main(input_args=None):
         help="Number of epochs training between each validation run",
     )
 
+    train_group.add_argument(
+        "--num_sanity_val_steps",
+        type=int,
+        default=2,
+        help="Number of sanity validation steps to run before training",
+    )
+
     # Evaluation options
     eval_group = parser.add_argument_group("Evaluation Options")
     eval_group.add_argument(
@@ -527,6 +534,7 @@ def main(input_args=None):
         callbacks=[val_checkpoint, latest_checkpoint],
         check_val_every_n_epoch=args.val_interval,
         precision=args.precision,
+        num_sanity_val_steps=args.num_sanity_val_steps,
     )
 
     # Only init once, on rank 0 only