diff --git a/src/ast/data_type.rs b/src/ast/data_type.rs
index 315d22b5..6b0ecd81 100644
--- a/src/ast/data_type.rs
+++ b/src/ast/data_type.rs
@@ -219,6 +219,10 @@ pub enum DataType {
     /// [hive]: https://docs.cloudera.com/cdw-runtime/cloud/impala-sql-reference/topics/impala-struct.html
     /// [bigquery]: https://cloud.google.com/bigquery/docs/reference/standard-sql/data-types#struct_type
     Struct(Vec<StructField>),
+    /// No type specified - only used with
+    /// [`SQLiteDialect`](crate::dialect::SQLiteDialect), from statements such
+    /// as `CREATE TABLE t1 (a)`.
+    Unspecified,
 }
 
 impl fmt::Display for DataType {
@@ -379,6 +383,7 @@ impl fmt::Display for DataType {
                     write!(f, "STRUCT")
                 }
             }
+            DataType::Unspecified => Ok(()),
         }
     }
 }
diff --git a/src/ast/ddl.rs b/src/ast/ddl.rs
index cd0cf179..cc3d5a53 100644
--- a/src/ast/ddl.rs
+++ b/src/ast/ddl.rs
@@ -586,7 +586,11 @@ pub struct ColumnDef {
 
 impl fmt::Display for ColumnDef {
     fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result {
-        write!(f, "{} {}", self.name, self.data_type)?;
+        if self.data_type == DataType::Unspecified {
+            write!(f, "{}", self.name)?;
+        } else {
+            write!(f, "{} {}", self.name, self.data_type)?;
+        }
         if let Some(collation) = &self.collation {
             write!(f, " COLLATE {collation}")?;
         }
diff --git a/src/dialect/mod.rs b/src/dialect/mod.rs
index 5167be24..8778948c 100644
--- a/src/dialect/mod.rs
+++ b/src/dialect/mod.rs
@@ -65,7 +65,9 @@ macro_rules! dialect_of {
 /// encapsulates the parsing differences between dialects.
 ///
 /// [`GenericDialect`] is the most permissive dialect, and parses the union of
-/// all the other dialects, when there is no ambiguity.
+/// all the other dialects, when there is no ambiguity. However, it does not
+/// currently allow `CREATE TABLE` statements without types specified for all
+/// columns; use [`SQLiteDialect`] if you require that.
 ///
 /// # Examples
 /// Most users create a [`Dialect`] directly, as shown on the [module
diff --git a/src/dialect/sqlite.rs b/src/dialect/sqlite.rs
index 0640466c..622fddee 100644
--- a/src/dialect/sqlite.rs
+++ b/src/dialect/sqlite.rs
@@ -16,6 +16,11 @@ use crate::keywords::Keyword;
 use crate::parser::{Parser, ParserError};
 
 /// A [`Dialect`] for [SQLite](https://www.sqlite.org)
+///
+/// This dialect allows columns in a
+/// [`CREATE TABLE`](https://sqlite.org/lang_createtable.html) statement with no
+/// type specified, as in `CREATE TABLE t1 (a)`. In the AST, these columns will
+/// have the data type [`Unspecified`](crate::ast::DataType::Unspecified).
 #[derive(Debug)]
 pub struct SQLiteDialect {}
 
diff --git a/src/parser/mod.rs b/src/parser/mod.rs
index 92f4cf02..28cabbb1 100644
--- a/src/parser/mod.rs
+++ b/src/parser/mod.rs
@@ -4272,7 +4272,11 @@ impl<'a> Parser<'a> {
 
     pub fn parse_column_def(&mut self) -> Result<ColumnDef, ParserError> {
         let name = self.parse_identifier()?;
-        let data_type = self.parse_data_type()?;
+        let data_type = if self.is_column_type_sqlite_unspecified() {
+            DataType::Unspecified
+        } else {
+            self.parse_data_type()?
+        };
         let mut collation = if self.parse_keyword(Keyword::COLLATE) {
             Some(self.parse_object_name()?)
         } else {
@@ -4308,6 +4312,29 @@ impl<'a> Parser<'a> {
         })
     }
 
+    fn is_column_type_sqlite_unspecified(&mut self) -> bool {
+        if dialect_of!(self is SQLiteDialect) {
+            match self.peek_token().token {
+                Token::Word(word) => matches!(
+                    word.keyword,
+                    Keyword::CONSTRAINT
+                        | Keyword::PRIMARY
+                        | Keyword::NOT
+                        | Keyword::UNIQUE
+                        | Keyword::CHECK
+                        | Keyword::DEFAULT
+                        | Keyword::COLLATE
+                        | Keyword::REFERENCES
+                        | Keyword::GENERATED
+                        | Keyword::AS
+                ),
+                _ => true, // e.g. comma immediately after column name
+            }
+        } else {
+            false
+        }
+    }
+
     pub fn parse_optional_column_option(&mut self) -> Result<Option<ColumnOption>, ParserError> {
         if self.parse_keywords(&[Keyword::CHARACTER, Keyword::SET]) {
             Ok(Some(ColumnOption::CharacterSet(self.parse_object_name()?)))
diff --git a/tests/sqlparser_sqlite.rs b/tests/sqlparser_sqlite.rs
index 0bf91d5b..8de862fc 100644
--- a/tests/sqlparser_sqlite.rs
+++ b/tests/sqlparser_sqlite.rs
@@ -221,6 +221,11 @@ fn parse_create_table_gencol() {
     sqlite_and_generic().verified_stmt("CREATE TABLE t1 (a INT, b INT AS (a * 2) STORED)");
 }
 
+#[test]
+fn parse_create_table_untyped() {
+    sqlite().verified_stmt("CREATE TABLE t1 (a, b AS (a * 2), c NOT NULL)");
+}
+
 #[test]
 fn test_placeholder() {
     // In postgres, this would be the absolute value operator '@' applied to the column 'xxx'