diff --git a/src/hz_sophia.erl b/src/hz_sophia.erl
index db47d22..b73a4c0 100644
--- a/src/hz_sophia.erl
+++ b/src/hz_sophia.erl
@@ -52,6 +52,8 @@ next_token(Tk, [C | _] = String) when ?IS_NUM(C) ->
     num_token(Tk, Tk, String, [], 0);
 next_token({tk, Row, Col}, [$#, C | Rest]) when ?IS_HEX(C) ->
     bytes_token({tk, Row, Col}, {tk, Row + 1, Col}, [C | Rest], "#", []);
+next_token({tk, Row, Col}, "\"" ++ Rest) ->
+    string_token({tk, Row, Col}, {tk, Row + 1, Col}, Rest, "\"", <<>>);
 next_token({tk, Row, Col}, [Char | Rest]) ->
     Token = {character, [Char], Char, Row, Col, Col},
     {ok, {Token, {tk, Row + 1, Col}, Rest}}.
@@ -101,6 +103,46 @@ reverse_combine_nibbles([D1], Acc) ->
 reverse_combine_nibbles([], Acc) ->
     Acc.
 
+string_token(Start, {tk, Row, Col}, [$\\, $x, A, B | Rest], SourceChars, Value) ->
+    case escape_hex_code(A, B) of
+        {ok, ByteVal} ->
+            string_token(Start, {tk, Row + 4, Col}, Rest, [B, A, $x, $\ | SourceChars], <<Value/binary, ByteVal>>);
+        error ->
+            {error, {invalid_escape_code, [$\\, $x, A, B], Row, Col}}
+    end;
+string_token(Start, {tk, Row, Col}, [$\\, C | Rest], SourceChars, Value) ->
+    case escape_char(C) of
+        {ok, ByteVal} ->
+            string_token(Start, {tk, Row + 2, Col}, Rest, [C, $\ | SourceChars], <<Value/binary, ByteVal>>);
+        error ->
+            {error, {invalid_escape_code, [C], Row, Col}}
+    end;
+string_token({tk, _, Start}, {tk, Row, End}, [$" | Rest], SourceChars, Value) ->
+    SourceStr = lists:reverse([$" | SourceChars]),
+    Token = {string, SourceStr, Value, Row, Start, End},
+    {ok, {Token, {tk, Row, End}, Rest}};
+string_token(Start, {tk, Row, Col}, [C | Rest], SourceChars, Value) ->
+    string_token(Start, {tk, Row + 1, Col}, Rest, [C | SourceChars], <<Value/binary, C>>).
+
+escape_hex_code(A, B) when ?IS_HEX(A), ?IS_HEX(B) ->
+    % As of writing this, the Sophia compiler will convert this byte from
+    % extended ASCII to unicode... But it really shouldn't. The literal parser
+    % does what the compiler should do.
+    Byte = convert_digit(A) * 16 + convert_digit(B),
+    {ok, Byte};
+escape_hex_code(_, _) ->
+    error.
+
+escape_char($b)  -> {ok, $\b};
+escape_char($e)  -> {ok, $\e};
+escape_char($f)  -> {ok, $\f};
+escape_char($n)  -> {ok, $\n};
+escape_char($r)  -> {ok, $\r};
+escape_char($t)  -> {ok, $\t};
+escape_char($v)  -> {ok, $\v};
+escape_char($")  -> {ok, $\"};
+escape_char($\\) -> {ok, $\\};
+escape_char(_)   -> error.
 
 %%% Sophia Literal Parser
 
@@ -121,8 +163,12 @@ reverse_combine_nibbles([], Acc) ->
 %%% pushdown automaton that we want.
 
 parse_expression(Type, Tk, String) ->
-    {ok, {Token, NewTk, NewString}} = next_token(Tk, String),
-    parse_expression2(Type, NewTk, NewString, Token).
+    case next_token(Tk, String) of
+        {ok, {Token, NewTk, NewString}} ->
+            parse_expression2(Type, NewTk, NewString, Token);
+        {error, Reason} ->
+            {error, Reason}
+    end.
 
 parse_expression2(Type, Tk, String, {integer, _, Value, Row, Start, End}) ->
     case Type of
@@ -146,7 +192,16 @@ parse_expression2(Type, Tk, String, {bytes, _, Value, Row, Start, End}) ->
         {_, _, unknown_type} ->
             {ok, {Result, Tk, String}};
         {O, N, _} ->
-            {error, {wrong_type, O, N, integer, Row, Start, End}}
+            {error, {wrong_type, O, N, {bytes, [Len]}, Row, Start, End}}
+    end;
+parse_expression2(Type, Tk, String, {string, _, Value, Row, Start, End}) ->
+    case Type of
+        {_, _, string} ->
+            {ok, {Value, Tk, String}};
+        {_, _, unknown_type} ->
+            {ok, {Value, Tk, String}};
+        {O, N, _} ->
+            {error, {wrong_type, O, N, string, Row, Start, End}}
     end;
 parse_expression2(Type, Tk, String, {character, "[", _, Row, Start, _}) ->
     parse_list(Type, Tk, String, Row, Start);
@@ -545,6 +600,10 @@ anon_types_test() ->
     % Bytes.
     check_parser("#DEAD000BEEF"),
     check_parser("#DE_AD0_00B_EEF"),
+    % Strings.
+    check_parser("\"hello world\""),
+    check_parser("\"  \\b\\e\\f\\n\\r\\t\\v\\\"\\\\  \""),
+    check_parser("\"\\x00\\x11\\x77\""),
     % List of integers.
     check_parser("[1, 2, 3]"),
     % List of lists.