diff --git a/rslib/src/search/parser.rs b/rslib/src/search/parser.rs
index 6937f02ab..3fb20c3e7 100644
--- a/rslib/src/search/parser.rs
+++ b/rslib/src/search/parser.rs
@@ -287,6 +287,7 @@ fn search_node_for_text_with_argument<'a>(
         "dupe" => parse_dupes(val)?,
         "prop" => parse_prop(val)?,
         "re" => SearchNode::Regex(unescape_quotes(val)),
+        "r" => SearchNode::UnqualifiedText(unescape_raw(val)),
         "nc" => SearchNode::NoCombining(unescape(val)?),
         "w" => SearchNode::WordBoundary(unescape(val)?),
         // anything else is a field search
@@ -419,6 +420,12 @@ fn parse_single_field<'a>(key: &'a str, val: &'a str) -> ParseResult<SearchNode<
             text: unescape_quotes(&val[3..]),
             is_re: true,
         }
+    } else if val.starts_with("r:") {
+        SearchNode::SingleField {
+            field: unescape(key)?,
+            text: unescape_raw(&val[2..]),
+            is_re: false,
+        }
     } else {
         SearchNode::SingleField {
             field: unescape(key)?,
@@ -437,6 +444,21 @@ fn unescape_quotes(s: &str) -> Cow<str> {
     }
 }
 
+/// Unescape quotes but escape wildcards and \s.
+fn unescape_raw(s: &str) -> Cow<str> {
+    lazy_static! {
+        static ref RE: Regex = Regex::new(r#"\\"?|\*|_"#).unwrap();
+    }
+    RE.replace_all(&s, |caps: &Captures| match &caps[0] {
+        r"\" => r"\\",
+        "\\\"" => "\"",
+        r"*" => r"\*",
+        r"_" => r"\_",
+        _ => unreachable!(),
+    })
+}
+
+/// Unescape chars with special meaning to the parser.
 fn unescape(txt: &str) -> ParseResult<Cow<str>> {
     if is_invalid_escape(txt) {
         Err(ParseError {})
@@ -611,12 +633,21 @@ mod test {
             vec![Search(Regex(r"\btest\%".into()))]
         );
 
+        // treat all chars as literals in raw searches
+        assert_eq!(parse(r"r:\*_"), parse(r"\\\*\_"));
+        assert_eq!(parse(r"field:r:\*_"), parse(r"field:\\\*\_"));
+
         // no exceptions for escaping "
         assert_eq!(
             parse(r#"re:te\"st"#)?,
             vec![Search(Regex(r#"te"st"#.into()))]
         );
         assert!(parse(r#"re:te"st"#).is_err());
+        assert_eq!(
+            parse(r#"r:te\"st"#)?,
+            vec![Search(UnqualifiedText(r#"te"st"#.into()))]
+        );
+        assert!(parse(r#"r:te"st"#).is_err());
 
         // spaces are optional if node separation is clear
         assert_eq!(parse(r#"a"b"(c)"#)?, parse("a b (c)")?);