diff --git a/vrobbler/apps/scrobbles/importers/webdav.py b/vrobbler/apps/scrobbles/importers/webdav.py
index 8d06044..9870ee2 100644
--- a/vrobbler/apps/scrobbles/importers/webdav.py
+++ b/vrobbler/apps/scrobbles/importers/webdav.py
@@ -507,7 +507,12 @@ def scan_webdav_for_ebird(webdav_client, user_id):
 
 
 def scan_webdav_for_scale(webdav_client, user_id):
-    """Download .csv files from WebDAV var/scale/ and queue imports for new files."""
+    """Download .csv files from WebDAV var/scale/ and queue imports for new files.
+
+    Because the scale CSV always has the same filename but grows by appending
+    rows, we detect changes by downloading the file and comparing its content
+    hash against the last completed import, rather than checking the filename.
+    """
     from scrobbles.models import ScaleCSVImport
     from scrobbles.tasks import process_scale_csv_import
 
@@ -530,28 +535,41 @@ def scan_webdav_for_scale(webdav_client, user_id):
         return 0
 
     new_imports = 0
-    already_imported = set(
-        ScaleCSVImport.objects.filter(user_id=user_id).values_list(
-            "original_filename", flat=True
-        )
-    )
 
     for fname in files:
         fname = os.path.basename(fname)
         if not fname.lower().endswith(".csv"):
             continue
-        if fname in already_imported:
-            logger.debug(f"Skipping already-imported {fname}")
-            continue
 
         tmp = tempfile.NamedTemporaryFile(delete=False, suffix=fname)
         try:
             webdav_client.download_sync(
                 remote_path=f"{scale_path}/{fname}", local_path=tmp.name
             )
+
+            new_hash = get_file_md5_hash(tmp.name)
+
+            last_import = (
+                ScaleCSVImport.objects.filter(
+                    user_id=user_id,
+                    original_filename=fname,
+                    processed_finished__isnull=False,
+                )
+                .order_by("-processed_finished")
+                .first()
+            )
+            if last_import and last_import.file_hash == new_hash:
+                logger.debug(
+                    "Scale CSV %s unchanged (hash match), skipping",
+                    fname,
+                    extra={"user_id": user_id, "hash": new_hash},
+                )
+                continue
+
             imp = ScaleCSVImport.objects.create(
                 user_id=user_id,
                 original_filename=fname,
+                file_hash=new_hash,
             )
             with open(tmp.name, "rb") as f:
                 imp.csv_file.save(fname, f, save=True)
diff --git a/vrobbler/apps/scrobbles/migrations/0088_scalecsvimport_file_hash.py b/vrobbler/apps/scrobbles/migrations/0088_scalecsvimport_file_hash.py
new file mode 100644
index 0000000..1b180de
--- /dev/null
+++ b/vrobbler/apps/scrobbles/migrations/0088_scalecsvimport_file_hash.py
@@ -0,0 +1,18 @@
+# Generated by Django 4.2.29 on 2026-05-28
+
+from django.db import migrations, models
+
+
+class Migration(migrations.Migration):
+
+    dependencies = [
+        ("scrobbles", "0087_koreaderimport_webdav_etag"),
+    ]
+
+    operations = [
+        migrations.AddField(
+            model_name="scalecsvimport",
+            name="file_hash",
+            field=models.CharField(blank=True, max_length=32, null=True),
+        ),
+    ]
diff --git a/vrobbler/apps/scrobbles/models.py b/vrobbler/apps/scrobbles/models.py
index 2509bfa..8ff6def 100644
--- a/vrobbler/apps/scrobbles/models.py
+++ b/vrobbler/apps/scrobbles/models.py
@@ -287,6 +287,7 @@ class ScaleCSVImport(BaseFileImportMixin):
 
     csv_file = models.FileField(upload_to=get_path, **BNULL)
     original_filename = models.CharField(max_length=255, **BNULL)
+    file_hash = models.CharField(max_length=32, **BNULL)
 
     def process(self, force=False):
         from scrobbles.importers.scale import import_scale_csv