diff --git a/src/lexer.cc b/src/lexer.cc
index bc3d3b8..a943c88 100644
--- a/src/lexer.cc
+++ b/src/lexer.cc
@@ -1,5 +1,7 @@
 #include <musique.hh>
 
+constexpr std::string_view Notes_Symbols = "abcedefgh";
+
 auto Lexer::next_token() -> Result<Token>
 {
 	while (consume_if(unicode::is_space)) {}
@@ -40,6 +42,51 @@ auto Lexer::next_token() -> Result<Token>
 		return { Token::Type::Numeric, finish(), token_location };
 	}
 
+	if (consume_if([](u32 ch) { return Notes_Symbols.find(ch) != std::string_view::npos; })) {
+		// chord declaration
+		constexpr u8 Expect_Number         = 0b01;
+		constexpr u8 Expect_Move           = 0b10;
+		constexpr u8 Expect_Number_Or_Move = 0b11;
+
+		auto current = Expect_Number;
+		std::string_view accepted_digits = "12357";
+		usize digit_cursor = 0;
+
+		for (;;) {
+			if ((current & Expect_Move) == Expect_Move
+			&& consume_if([](u32 c) { return c == ',' || c == '\''; })
+			) {
+				current = Expect_Number;
+				continue;
+			}
+
+			if ((current & Expect_Number) == Expect_Number) {
+				bool found = false;
+				for (; digit_cursor < accepted_digits.size(); ++digit_cursor) {
+					if (consume_if([&](u32 c) { return u32(accepted_digits[digit_cursor]) == c; })) {
+						found = true;
+						break;
+					}
+				}
+
+				if (found) {
+					current = digit_cursor < accepted_digits.size()
+						? Expect_Number_Or_Move
+						: Expect_Move;
+					continue;
+				}
+			}
+
+			break;
+		}
+
+		if (unicode::is_letter(peek())) {
+			assert(false && "symbols are not implemented yet");
+		}
+
+		return { Token::Type::Chord, finish(), token_location };
+	}
+
 	return errors::unrecognized_character(peek(), token_location);
 }
 
diff --git a/src/musique.hh b/src/musique.hh
index 89eebe3..00d57e4 100644
--- a/src/musique.hh
+++ b/src/musique.hh
@@ -113,6 +113,7 @@ namespace unicode
 
 	bool is_digit(u32 digit);
 	bool is_space(u32 space);
+	bool is_letter(u32 letter);
 }
 
 namespace utf8
@@ -183,7 +184,7 @@ struct Lexer
 	// Finds next rune in source and returns it, advancing the string
 	auto consume() -> u32;
 
-	inline auto consume_if(auto test) -> u32
+	inline auto consume_if(auto test) -> bool
 	{
 		return test(peek()) && (consume(), true);
 	}
diff --git a/src/tests/lex.cc b/src/tests/lex.cc
index 2319d11..4cfed50 100644
--- a/src/tests/lex.cc
+++ b/src/tests/lex.cc
@@ -27,9 +27,12 @@ static void expect_token_type_and_value(
 {
 	Lexer lexer{source};
 	auto result = lexer.next_token();
-	expect(result.has_value() >> fatal, sl) << "have not parsed any tokens";
-	expect(eq(under(result->type), under(expected_type)), sl) << "different token type then expected";
-	expect(eq(result->source, expected), sl) << "tokenized source is not equal to original";
+	expect(result.has_value(), sl) << "have not parsed any tokens";
+
+	if (result.has_value()) {
+		expect(eq(under(result->type), under(expected_type)), sl) << "different token type then expected";
+		expect(eq(result->source, expected), sl) << "tokenized source is not equal to original";
+	}
 }
 
 static void expect_token_type_and_value(
@@ -86,4 +89,13 @@ suite lexer_test = [] {
 		expect_token_type_and_location(Token::Type::Numeric, "\n123", Location::at(2, 1));
 		expect_token_type_and_location(Token::Type::Numeric, "\n  123", Location::at(2, 3));
 	};
+
+	"Chord literals"_test = [] {
+		expect_token_type_and_value(Token::Type::Chord, "c1");
+		expect_token_type_and_value(Token::Type::Chord, "d1257");
+		expect_token_type_and_value(Token::Type::Chord, "e1'");
+		expect_token_type_and_value(Token::Type::Chord, "g127");
+		expect_token_type_and_value(Token::Type::Chord, "f1'2'3'5'7'");
+		expect_token_type_and_value(Token::Type::Chord, "b1,2,5,7,");
+	};
 };
diff --git a/src/unicode.cc b/src/unicode.cc
index 3c6dcaa..82b851d 100644
--- a/src/unicode.cc
+++ b/src/unicode.cc
@@ -97,3 +97,9 @@ bool unicode::is_space(u32 space)
 	}
 	return false;
 }
+
+bool unicode::is_letter(u32 letter)
+{
+	// TODO Unicode letters handling
+	return std::isalpha(letter);
+}