diff --git a/src/grammar.lalrpop b/src/grammar.lalrpop
index 174ec6326e..18c0390dcf 100644
--- a/src/grammar.lalrpop
+++ b/src/grammar.lalrpop
@@ -181,6 +181,7 @@ UOp: UnaryOp<RichTerm> = {
     "head" => UnaryOp::ListHead(),
     "tail" => UnaryOp::ListTail(),
     "length" => UnaryOp::ListLength(),
+    "fieldsOf" => UnaryOp::FieldsOf(),
 };
 
 switch_case: (Ident, RichTerm) = {
@@ -354,6 +355,7 @@ extern {
         "head" => Token::Head,
         "tail" => Token::Tail,
         "length" => Token::Length,
+        "fieldsOf" => Token::FieldsOf,
 
         "hasField" => Token::HasField,
         "map" => Token::Map,
diff --git a/src/operation.rs b/src/operation.rs
index f5d84f3fdb..1887cb7d7d 100644
--- a/src/operation.rs
+++ b/src/operation.rs
@@ -328,6 +328,21 @@ fn process_unary_operation(
                 ))
             }
         }
+        UnaryOp::FieldsOf() => {
+            if let Term::Record(map) = *t {
+                let mut fields: Vec<String> = map.keys().map(|Ident(id)| id.clone()).collect();
+                fields.sort();
+                let terms = fields.into_iter().map(|id| Term::Str(id).into()).collect();
+                Ok(Closure::atomic_closure(Term::List(terms).into()))
+            } else {
+                Err(EvalError::TypeError(
+                    String::from("Record"),
+                    String::from("fieldsOf"),
+                    arg_pos,
+                    RichTerm { term: t, pos },
+                ))
+            }
+        }
         UnaryOp::MapRec(f) => {
             if let Term::Record(rec) = *t {
                 let f_as_var = f.body.closurize(&mut env, f.env);
diff --git a/src/parser/lexer.rs b/src/parser/lexer.rs
index 7c454f5605..e139c5ae9b 100644
--- a/src/parser/lexer.rs
+++ b/src/parser/lexer.rs
@@ -107,6 +107,7 @@ pub enum Token<'input> {
     Head,
     Tail,
     Length,
+    FieldsOf,
 
     Unwrap,
     HasField,
@@ -219,6 +220,7 @@ impl<'input> fmt::Display for Token<'input> {
             Token::Head => "head",
             Token::Tail => "tail",
             Token::Length => "length",
+            Token::FieldsOf => "fieldsOf",
 
             Token::HasField => "hasField",
             Token::Map => "map",
@@ -559,6 +561,7 @@ impl<'input> Lexer<'input> {
             "map" => Token::Map,
             "elemAt" => Token::ElemAt,
             "merge" => Token::Merge,
+            "fieldsOf" => Token::FieldsOf,
             ty @ "Dyn" | ty @ "Num" | ty @ "Bool" | ty @ "Str" | ty @ "List" => Token::Type(ty),
             id => Token::Identifier(id),
         };
diff --git a/src/program.rs b/src/program.rs
index 5c7b91c185..8e981c103c 100644
--- a/src/program.rs
+++ b/src/program.rs
@@ -1253,4 +1253,15 @@ Assume(#alwaysTrue -> #alwaysFalse, not ) true
         assert_npeq!("{ a = \"a\"; b = true }", "{ a = true; b = \"a\"}");
         assert_npeq!("{ a = { a = true } }", "{a = { a = { a = true } } }");
     }
+
+    #[test]
+    fn fields_of() {
+        assert_peq!("fieldsOf {}", "[]");
+        assert_peq!("fieldsOf {a = 1; b = 2; c = 3}", "[\"a\", \"b\", \"c\"]");
+        assert_peq!("fieldsOf {aAa = 1; Zzz = 2;}", "[\"Zzz\", \"aAa\"]");
+        assert_peq!(
+            "fieldsOf {foo = {bar = 0}; baz = Default(true)}",
+            "[\"baz\", \"foo\"]"
+        );
+    }
 }
diff --git a/src/term.rs b/src/term.rs
index 37b7f46978..e684d3d8e1 100644
--- a/src/term.rs
+++ b/src/term.rs
@@ -500,6 +500,9 @@ pub enum UnaryOp<CapturedTerm> {
     /// string accumulator, the remaining chunks to be evaluated, and is applied to the current
     /// chunk being evaluated.
     ChunksConcat(String, Vec<StrChunk<CapturedTerm>>),
+
+    /// Return the names of the fields of a record as a string list.
+    FieldsOf(),
 }
 
 impl<Ty> UnaryOp<Ty> {
@@ -559,6 +562,8 @@ impl<Ty> UnaryOp<Ty> {
                     })
                     .collect(),
             ),
+
+            FieldsOf() => FieldsOf(),
         }
     }
 }
diff --git a/src/typecheck.rs b/src/typecheck.rs
index 325dca19dc..4071cae5cd 100644
--- a/src/typecheck.rs
+++ b/src/typecheck.rs
@@ -1379,6 +1379,13 @@ pub fn get_uop_type(
         )),
         // This should not happen, as ChunksConcat() is only produced during evaluation.
         UnaryOp::ChunksConcat(_, _) => panic!("cannot type ChunksConcat()"),
+        // forall rows. { rows } -> List
+        UnaryOp::FieldsOf() => TypeWrapper::Concrete(AbsType::arrow(
+            Box::new(TypeWrapper::Concrete(AbsType::StaticRecord(Box::new(
+                TypeWrapper::Ptr(new_var(state.table)),
+            )))),
+            Box::new(TypeWrapper::Concrete(AbsType::List())),
+        )),
     })
 }