diff --git a/parsers/caves.py b/parsers/caves.py
index 7f93cb3..4a4584c 100644
--- a/parsers/caves.py
+++ b/parsers/caves.py
@@ -920,7 +920,8 @@ def read_cave(filename, mvf=None, cave=None):
 
     slug = check_slug(cave.areacode, cave.kataster_number, cave.unofficial_number, cave.url) #NB cave.slug is not a field on Cave
                
-    # Thsi whole way of doing entrances can be replaced by simply knowing the entrance_data filename what the cave is. to do.
+    # Thsi whole way of doing entrances can be replaced by simply knowing the entrance_data filename what the cave is. TO DO.
+    # This also fails if there are more entance_data files for a cave tahn the cave_data file knows about.
     entrances = getXML(cavecontents, "entrance", context=context)
     #do_entrances()
     if not entrances or len(entrances) < 1:
@@ -1015,6 +1016,15 @@ def readcaves():
                 if filename.endswith(".html"):
                     read_cave(filename, mvf)
 
+    # check entrances all have caves associated with them
+    for e in Entrance.objects.all():
+        if not e.firstcave():
+            message = f'  ! No cave associated with entrance {e.slug} in {ENTRANCEDESCRIPTIONS}/{e.filename}. Deleting the Entrance object.'
+            DataIssue.objects.create(parser="entrances", message=message, url=f"/{e.slug}")
+            print(message)
+            e.delete()
+            
+    
     print(" - Setting up all the variously useful alias names")
     GetCaveLookup()