diff --git a/everyvoice/wizard/basic.py b/everyvoice/wizard/basic.py
index 60ac793b..2b3090d3 100644
--- a/everyvoice/wizard/basic.py
+++ b/everyvoice/wizard/basic.py
@@ -1,4 +1,6 @@
 import json
+import sys
+import wave
 from pathlib import Path
 
 import questionary
@@ -199,6 +201,42 @@ class ConfigFormatStep(Step):
     # ConfigFormatStep writes the results to disk and exits, so it's not reversible.
     REVERSIBLE = False
 
+    def validate_wav_channels(self, wavs_dir: Path):
+        """Validate the number of channels in WAV files."""
+        wav_files = list(wavs_dir.rglob("*.wav"))  # Recursively find all WAV files.
+
+        # if not wav_files:
+        #    rich_print(
+        #        f"[red]No WAV files found in the directory '{wavs_dir}'. Please check your files.[/red]"
+        #    )
+        #    sys.exit(1)  # Exit if no WAV files are found.
+
+        rich_print(
+            f"Validating the number of channels in WAV files located at '{wavs_dir}'..."
+        )
+        for wav_file in wav_files:
+            try:
+                with wave.open(str(wav_file), "rb") as wav:
+                    num_channels = wav.getnchannels()
+                    if num_channels > 2:
+                        rich_print(
+                            f"[red]Error: The file '{wav_file}' has {num_channels} channels. "
+                            "Only single-channel or two-channel WAV files are supported. "
+                            "Please correct this and try again.[/red]"
+                        )
+                        sys.exit(
+                            1
+                        )  # Exit gracefully if a file has more than 2 channels.
+            except wave.Error as e:
+                rich_print(
+                    f"[red]Error reading WAV file '{wav_file}': {e}. Please check the file.[/red]"
+                )
+                sys.exit(1)  # Gracefully exit if the file cannot be read.
+
+        rich_print(
+            "[green]All WAV files have valid channel counts (1 or 2 channels).[/green]"
+        )
+
     def prompt(self):
         return get_response_from_menu_prompt(
             "Which format would you like to output the configuration to?",
@@ -220,7 +258,7 @@ def effect(self):
         preprocessed_dir.absolute().mkdir(parents=True, exist_ok=True)
         # used in configs
         preprocessed_dir_relative_to_configs = Path("..") / "preprocessed"
-        # log dir
+        # log dirself.state
         log_dir = output_path / "logs_and_checkpoints"
         log_dir.absolute().mkdir(parents=True, exist_ok=True)
         log_dir_relative_to_configs = Path("..") / "logs_and_checkpoints"
@@ -236,6 +274,13 @@ def effect(self):
         )  # TODO: this should be fixed by https://github.com/EveryVoiceTTS/EveryVoice/issues/359
         for dataset in [key for key in self.state.keys() if key.startswith("dataset_")]:
             dataset_state = self.state[dataset]
+
+            # Validate WAV Channels
+            wavs_dir = Path(dataset_state[StepNames.wavs_dir_step]).expanduser()
+            self.validate_wav_channels(
+                wavs_dir
+            )  # Call the new validation function here
+
             # Add Cleaners
             # TODO: these should really be dataset-specific cleaners, not global cleaners
             # so this should be fixed by https://github.com/EveryVoiceTTS/EveryVoice/issues/359