diff --git a/CHANGELOG.rst b/CHANGELOG.rst
index cf4db41092..0df9b8fb64 100644
--- a/CHANGELOG.rst
+++ b/CHANGELOG.rst
@@ -8,6 +8,10 @@ Next release
   ``licensedcode-data``.
   https://github.com/aboutcode-org/scancode-toolkit/pull/5056
 
+- Fix a regression in PDF text extraction that silently stopped after the
+  first page, so copyright/license detection missed text on later pages.
+  https://github.com/aboutcode-org/scancode-toolkit/pull/5260
+
 - Improve copyright detection for statements with parens or trailing "authors"
 
 
diff --git a/src/textcode/pdf.py b/src/textcode/pdf.py
index 4e90c866f8..4beb11aea6 100644
--- a/src/textcode/pdf.py
+++ b/src/textcode/pdf.py
@@ -22,8 +22,9 @@
 
 def get_text_lines(location, max_pages=5):
     """
-    Return a list of unicode text lines extracted from a pdf file at
-    `location`. May raise exceptions. Extract up to `max_pages` pages.
+    Return a list of text lines, as bytes, extracted from a pdf file at
+    `location`. May raise exceptions. Extract up to `max_pages` pages, or
+    all pages if `max_pages` is 0.
     """
     extracted_text = BytesIO()
     laparams = LAParams()
@@ -43,5 +44,5 @@ def get_text_lines(location, max_pages=5):
                 interpreter.process_page(page)
                 if max_pages and page_num == max_pages:
                     break
-                extracted_text.seek(0)
-                return extracted_text.readlines()
+            extracted_text.seek(0)
+            return extracted_text.readlines()
diff --git a/tests/licensedcode/data/datadriven/lic4/should_detect_something_5.pdf.yml b/tests/licensedcode/data/datadriven/lic4/should_detect_something_5.pdf.yml
index 8aa82b1c4c..39c8442d82 100644
--- a/tests/licensedcode/data/datadriven/lic4/should_detect_something_5.pdf.yml
+++ b/tests/licensedcode/data/datadriven/lic4/should_detect_something_5.pdf.yml
@@ -1,4 +1,2 @@
 license_expressions:
   - sun-sissl-1.1
-  - proprietary-license
-  - cpal-1.0
diff --git a/tests/textcode/data/pdf/multi_page.pdf b/tests/textcode/data/pdf/multi_page.pdf
new file mode 100644
index 0000000000..d369663e18
Binary files /dev/null and b/tests/textcode/data/pdf/multi_page.pdf differ
diff --git a/tests/textcode/test_pdf.py b/tests/textcode/test_pdf.py
index 7941953568..cc7b2d05da 100644
--- a/tests/textcode/test_pdf.py
+++ b/tests/textcode/test_pdf.py
@@ -51,6 +51,33 @@ def get_text(location):
 
         assert result == expected
 
+    def test_get_text_lines_extracts_all_pages_up_to_max_pages(self):
+        # regression test: text extraction must not stop after the first page
+        test_file = self.get_test_loc('pdf/multi_page.pdf')
+        result = pdf.get_text_lines(test_file)
+        text = b''.join(result)
+        assert b'This is page 1 of a multi-page test document.' in text
+        assert b'This notice is on page 2.' in text
+        assert b'The last page 5 of this test document.' in text
+        # the default max_pages=5 must still be honored
+        assert b'Page 6' not in text
+
+    def test_get_text_lines_returns_lines_when_max_pages_is_one(self):
+        # regression test: reaching max_pages must not return None
+        test_file = self.get_test_loc('pdf/multi_page.pdf')
+        result = pdf.get_text_lines(test_file, max_pages=1)
+        assert result
+        text = b''.join(result)
+        assert b'This is page 1 of a multi-page test document.' in text
+        assert b'page 2' not in text
+
+    def test_get_text_lines_extracts_all_pages_when_max_pages_is_zero(self):
+        test_file = self.get_test_loc('pdf/multi_page.pdf')
+        result = pdf.get_text_lines(test_file, max_pages=0)
+        text = b''.join(result)
+        assert b'Page 6 is beyond the default max_pages limit.' in text
+        assert b'Page 7 is also beyond the default max_pages limit.' in text
+
     def test_pdfminer_can_parse_faulty_broadcom_doc(self):
         # test for https://github.com/euske/pdfminer/issues/118
         test_file = self.get_test_loc('pdf/pdfminer_bug_118/faulty.pdf')