diff --git a/CHANGELOG b/CHANGELOG
index eb5c381a..41a4c2af 100644
--- a/CHANGELOG
+++ b/CHANGELOG
@@ -9,6 +9,7 @@ Enhancements
 Bug Fixes
 
 * Fix parsing of backticks (issue588).
+* Fix parsing of scientific number (issue399).
 
 
 Release 0.4.1 (Oct 08, 2020)
diff --git a/sqlparse/keywords.py b/sqlparse/keywords.py
index 26dd3b38..68506282 100644
--- a/sqlparse/keywords.py
+++ b/sqlparse/keywords.py
@@ -62,7 +62,7 @@ def is_keyword(value):
         (r'(?<=\.)[A-ZÀ-Ü]\w*', tokens.Name),  # .'Name'
         (r'[A-ZÀ-Ü]\w*(?=\()', tokens.Name),  # side effect: change kw to func
         (r'-?0x[\dA-F]+', tokens.Number.Hexadecimal),
-        (r'-?\d*(\.\d+)?E-?\d+', tokens.Number.Float),
+        (r'-?\d+(\.\d+)?E-?\d+', tokens.Number.Float),
         (r'(?![_A-ZÀ-Ü])-?(\d+(\.\d*)|\.\d+)(?![_A-ZÀ-Ü])',
          tokens.Number.Float),
         (r'(?![_A-ZÀ-Ü])-?\d+(?![_A-ZÀ-Ü])', tokens.Number.Integer),
diff --git a/tests/test_parse.py b/tests/test_parse.py
index 5fe6ed23..513b4be9 100644
--- a/tests/test_parse.py
+++ b/tests/test_parse.py
@@ -188,11 +188,16 @@ def test_placeholder(ph):
     assert p[0].ttype is T.Name.Placeholder
 
 
-@pytest.mark.parametrize('num', ['6.67428E-8', '1.988e33', '1e-12'])
-def test_scientific_numbers(num):
+@pytest.mark.parametrize('num, expected', [
+    ('6.67428E-8', T.Number.Float),
+    ('1.988e33', T.Number.Float),
+    ('1e-12', T.Number.Float),
+    ('e1', None),
+])
+def test_scientific_numbers(num, expected):
     p = sqlparse.parse(num)[0].tokens
     assert len(p) == 1
-    assert p[0].ttype is T.Number.Float
+    assert p[0].ttype is expected
 
 
 def test_single_quotes_are_strings():
@@ -336,7 +341,8 @@ def test_pprint():
         "|  |  `- 0 Name 'd0'",
         "|  |- 10 Punctuation ','",
         "|  |- 11 Whitespace ' '",
-        "|  `- 12 Float 'e0'",
+        "|  `- 12 Identifier 'e0'",
+        "|     `- 0 Name 'e0'",
         "|- 3 Whitespace ' '",
         "|- 4 Keyword 'from'",
         "|- 5 Whitespace ' '",