diff --git a/bnd/pipeline/pyaldata.py b/bnd/pipeline/pyaldata.py
index 7f271b3..b0ee052 100644
--- a/bnd/pipeline/pyaldata.py
+++ b/bnd/pipeline/pyaldata.py
@@ -736,9 +736,6 @@ def expand_dim_in_single_bin_trials(self, column_subset="_spikes") -> None:
         column_subset :
             String expression to look for in columns to be expanded. Defaults to 'spikes_'
 
-        Returns
-        -------
-
         """
 
         def _expand_dim_in_single_bin_trial(value):
@@ -752,13 +749,24 @@ def _expand_dim_in_single_bin_trial(value):
 
         return
 
+    def drop_empty_states_at_end(self) -> None:
+        """
+        Drop last column if all spike fields are empty
+
+        """
+        spike_cols = [col for col in self.pyaldata_df.columns if col.endswith("_spikes")]
+        final_state = self.pyaldata_df.iloc[-1]
+        all_zero = all(final_state[spike_col].size == 0 for spike_col in spike_cols)
+
+        if all_zero:
+            self.pyaldata_df.drop(self.pyaldata_df.index[-1], inplace=True)
+
+        return
+
     def run_conversion(self):
         """
         Main routine for pyaldata conversion
 
-        Returns
-        -------
-
         """
 
         # Define all the necessary columns
@@ -794,6 +802,9 @@ def run_conversion(self):
         # Expand dimensions
         self.expand_dim_in_single_bin_trials()
 
+        # Drop empty states at the end
+        self.drop_empty_states_at_end()
+
         logger.info("Session converted to pyaldata format")
 
         return