diff --git a/fedireads/goodreads_import.py b/fedireads/goodreads_import.py
index 1333f9cc..adcbad3c 100644
--- a/fedireads/goodreads_import.py
+++ b/fedireads/goodreads_import.py
@@ -2,9 +2,9 @@
 import re
 import csv
 import itertools
-from requests import HTTPError
 
 from fedireads import books_manager
+from fedireads.models import Edition
 
 
 # Mapping goodreads -> fedireads shelf titles.
@@ -42,13 +42,7 @@ class GoodreadsCsv:
 
     def __iter__(self):
         for line in itertools.islice(self.reader, MAX_ENTRIES):
-            entry = GoodreadsItem(line)
-            try:
-                entry.resolve()
-            except HTTPError:
-                pass
-            yield entry
-
+            yield GoodreadsItem(line)
 
 class GoodreadsItem:
     ''' a processed line in a goodreads csv '''
@@ -56,22 +50,27 @@ class GoodreadsItem:
         self.line = line
         self.book = None
 
-
     def resolve(self):
         ''' try various ways to lookup a book '''
-        self.book = self.get_book_from_isbn()
-        if not self.book:
-            self.book = self.get_book_from_title_author()
+        self.book = (
+            self.get_book_from_db_isbn() or
+            self.get_book_from_isbn() or
+            self.get_book_from_title_author()
+        )
 
+    def get_book_from_db_isbn(self):
+        ''' see if we already know about the book '''
+        try:
+            return Edition.objects.get(isbn=self.isbn)
+        except Edition.DoesNotExist:
+            return None
 
     def get_book_from_isbn(self):
         ''' search by isbn '''
-        isbn = unquote_string(self.line['ISBN13'])
-        search_results = books_manager.search(isbn)
+        search_results = books_manager.search(self.isbn)
         if search_results:
             return books_manager.get_or_create_book(search_results[0].key)
 
-
     def get_book_from_title_author(self):
         ''' search by title and author '''
         search_term = construct_search_term(
@@ -82,6 +81,10 @@ class GoodreadsItem:
         if search_results:
             return books_manager.get_or_create_book(search_results[0].key)
 
+    @property
+    def isbn(self):
+        return unquote_string(self.line['ISBN13'])
+
     @property
     def shelf(self):
         ''' the goodreads shelf field '''
diff --git a/fedireads/view_actions.py b/fedireads/view_actions.py
index 7ccf4888..c3134441 100644
--- a/fedireads/view_actions.py
+++ b/fedireads/view_actions.py
@@ -2,6 +2,7 @@
 from io import BytesIO, TextIOWrapper
 import re
 from PIL import Image
+from requests import HTTPError
 
 from django.contrib.auth import authenticate, login, logout
 from django.contrib.auth.decorators import login_required
@@ -424,6 +425,10 @@ def import_data(request):
         for item in GoodreadsCsv(TextIOWrapper(
                 request.FILES['csv_file'],
                 encoding=request.encoding)):
+            try:
+                item.resolve()
+            except HttpError:
+                pass
             if item.book:
                 results.append(item)
                 if item.rating or item.review: