diff --git a/graphql-c_parser/ext/graphql_c_parser_ext/lexer.c b/graphql-c_parser/ext/graphql_c_parser_ext/lexer.c
index 1433f8993b..e0f0c1f2d5 100644
--- a/graphql-c_parser/ext/graphql_c_parser_ext/lexer.c
+++ b/graphql-c_parser/ext/graphql_c_parser_ext/lexer.c
@@ -868,7 +868,7 @@ void emit(TokenType tt, char *ts, char *te, Meta *meta) {
 		token_sym = ID2SYM(rb_intern("STRING"));
 		quotes_length = 3;
 		token_content = rb_utf8_str_new(ts + quotes_length, (te - ts - (2 * quotes_length)));
-		line_incr = FIX2INT(rb_funcall(token_content, rb_intern("count"), 1, rb_str_new_cstr("\n")));
+		line_incr = FIX2INT(rb_funcall(token_content, rb_intern("count"), 1, rb_utf8_str_new_cstr("\n")));
 		break;
 		case STRING:
 		// This is used only by the parser, this is never reached
@@ -1976,12 +1976,12 @@ VALUE tokenize(VALUE query_rbstr) {
 
 
 #define SETUP_STATIC_TOKEN_VARIABLE(token_name, token_content) \
-GraphQLTokenString##token_name = rb_str_new_cstr(token_content); \
+GraphQLTokenString##token_name = rb_utf8_str_new_cstr(token_content); \
 rb_funcall(GraphQLTokenString##token_name, rb_intern("-@"), 0); \
 rb_global_variable(&GraphQLTokenString##token_name); \
 
 #define SETUP_STATIC_STRING(var_name, str_content) \
-var_name = rb_str_new_cstr(str_content); \
+var_name = rb_utf8_str_new_cstr(str_content); \
 rb_global_variable(&var_name); \
 rb_str_freeze(var_name); \
 
diff --git a/graphql-c_parser/ext/graphql_c_parser_ext/lexer.rl b/graphql-c_parser/ext/graphql_c_parser_ext/lexer.rl
index b3f36024ea..edadb4b4d1 100644
--- a/graphql-c_parser/ext/graphql_c_parser_ext/lexer.rl
+++ b/graphql-c_parser/ext/graphql_c_parser_ext/lexer.rl
@@ -282,7 +282,7 @@ void emit(TokenType tt, char *ts, char *te, Meta *meta) {
       token_sym = ID2SYM(rb_intern("STRING"));
       quotes_length = 3;
       token_content = rb_utf8_str_new(ts + quotes_length, (te - ts - (2 * quotes_length)));
-      line_incr = FIX2INT(rb_funcall(token_content, rb_intern("count"), 1, rb_str_new_cstr("\n")));
+      line_incr = FIX2INT(rb_funcall(token_content, rb_intern("count"), 1, rb_utf8_str_new_cstr("\n")));
       break;
     case STRING:
       // This is used only by the parser, this is never reached
@@ -359,12 +359,12 @@ VALUE tokenize(VALUE query_rbstr) {
 
 
 #define SETUP_STATIC_TOKEN_VARIABLE(token_name, token_content) \
-  GraphQLTokenString##token_name = rb_str_new_cstr(token_content); \
+  GraphQLTokenString##token_name = rb_utf8_str_new_cstr(token_content); \
   rb_funcall(GraphQLTokenString##token_name, rb_intern("-@"), 0); \
   rb_global_variable(&GraphQLTokenString##token_name); \
 
 #define SETUP_STATIC_STRING(var_name, str_content) \
-  var_name = rb_str_new_cstr(str_content); \
+  var_name = rb_utf8_str_new_cstr(str_content); \
   rb_global_variable(&var_name); \
   rb_str_freeze(var_name); \
 
diff --git a/spec/graphql/language/lexer_examples.rb b/spec/graphql/language/lexer_examples.rb
index 02a78839db..ad264e991b 100644
--- a/spec/graphql/language/lexer_examples.rb
+++ b/spec/graphql/language/lexer_examples.rb
@@ -58,6 +58,12 @@ def self.included(child_mod)
           assert_equal :BAD_UNICODE_ESCAPE, subject.tokenize(string).first.name
         end
 
+        it "makes utf-8 arguments named type" do
+          str = "{ a(type: 1) }"
+          tokens = subject.tokenize(str)
+          assert_equal Encoding::UTF_8, tokens[2].value.encoding
+        end
+
         it "makes utf-8 comments" do
           tokens = subject.tokenize("# 不要!\n{")
           comment_token = tokens.first.prev_token