diff --git a/lib/makeup/lexers/elixir_lexer.ex b/lib/makeup/lexers/elixir_lexer.ex
index 4d3404e..48e17b6 100644
--- a/lib/makeup/lexers/elixir_lexer.ex
+++ b/lib/makeup/lexers/elixir_lexer.ex
@@ -25,7 +25,7 @@ defmodule Makeup.Lexers.ElixirLexer do
   # TODO: check we're following this convention
   # NOTE: if Elixir had a good static type system it would help us do the right thing here.
 
-  whitespace = ascii_string([?\r, ?\s, ?\n, ?\f], min: 1) |> token(:whitespace)
+  whitespace = ascii_string([?\r, ?\s, ?\n, ?\t], min: 1) |> token(:whitespace)
 
   newlines =
     optional(ascii_string([?\s, ?\t, ?\r], min: 1))
diff --git a/test/makeup/lexers/elixir_lexer/elixir_lexer_tokenizer_test.exs b/test/makeup/lexers/elixir_lexer/elixir_lexer_tokenizer_test.exs
index 3ac5c4d..f93d185 100644
--- a/test/makeup/lexers/elixir_lexer/elixir_lexer_tokenizer_test.exs
+++ b/test/makeup/lexers/elixir_lexer/elixir_lexer_tokenizer_test.exs
@@ -2,8 +2,12 @@ defmodule ElixirLexerTokenizerTestSnippet do
   use ExUnit.Case, async: false
   import Makeup.Lexers.ElixirLexer.Testing, only: [lex: 1]
 
-  test "?\\f is recognized as whitespace character" do
-    assert lex("\f") == [{:whitespace, %{}, "\f"}]
+  test "whitespace" do
+    assert lex(" ") == [{:whitespace, %{}, " "}]
+    assert lex("\s") == [{:whitespace, %{}, "\s"}]
+    assert lex("\n") == [{:whitespace, %{}, "\n"}]
+    assert lex("\t") == [{:whitespace, %{}, "\t"}]
+    assert lex("\r") == [{:whitespace, %{}, "\r"}]
   end
 
   test "builtins" do