diff --git a/lib/floki.ex b/lib/floki.ex
index 4135b150..2892d553 100644
--- a/lib/floki.ex
+++ b/lib/floki.ex
@@ -790,4 +790,19 @@ defmodule Floki do
   def filter_out(elements, selector) do
     FilterOut.filter_out(elements, selector)
   end
+
+  @doc """
+  Escapes a string for use as a CSS identifier.
+
+  ## Examples
+
+      iex> Floki.css_escape("hello world")
+      "hello\\\\ world"
+
+      iex> Floki.css_escape("-123")
+      "-\\\\31 23"
+
+  """
+  @spec css_escape(String.t()) :: String.t()
+  def css_escape(value), do: Floki.CSSEscape.escape(value)
 end
diff --git a/lib/floki/css_escape.ex b/lib/floki/css_escape.ex
new file mode 100644
index 00000000..7eb6036d
--- /dev/null
+++ b/lib/floki/css_escape.ex
@@ -0,0 +1,77 @@
+defmodule Floki.CSSEscape do
+  @moduledoc false
+
+  # This is a direct translation of
+  # https://github.com/mathiasbynens/CSS.escape/blob/master/css.escape.js
+  # into Elixir.
+
+  @doc """
+  Escapes a string for use as a CSS identifier.
+
+  ## Examples
+
+      iex> Floki.CSSEscape.escape("hello world")
+      "hello\\\\ world"
+
+      iex> Floki.CSSEscape.escape("-123")
+      "-\\\\31 23"
+
+  """
+  @spec escape(String.t()) :: String.t()
+  def escape(value) when is_binary(value) do
+    value
+    |> String.to_charlist()
+    |> escape_chars()
+    |> IO.iodata_to_binary()
+  end
+
+  def escape(_), do: raise(ArgumentError, "CSS.escape requires a string argument")
+
+  defp escape_chars(chars) do
+    case chars do
+      # If the character is the first character and is a `-` (U+002D), and
+      # there is no second character, […]
+      [?- | []] -> ["\\-"]
+      _ -> do_escape_chars(chars, 0, [])
+    end
+  end
+
+  defp do_escape_chars([], _, acc), do: Enum.reverse(acc)
+
+  defp do_escape_chars([char | rest], index, acc) do
+    escaped =
+      cond do
+        # If the character is NULL (U+0000), then the REPLACEMENT CHARACTER
+        # (U+FFFD).
+        char == 0 ->
+          <<0xFFFD::utf8>>
+
+        # If the character is in the range [\1-\1F] (U+0001 to U+001F) or is
+        # U+007F,
+        # if the character is the first character and is in the range [0-9]
+        # (U+0030 to U+0039),
+        # if the character is the second character and is in the range [0-9]
+        # (U+0030 to U+0039) and the first character is a `-` (U+002D),
+        char in 0x0001..0x001F or char == 0x007F or
+          (index == 0 and char in ?0..?9) or
+            (index == 1 and char in ?0..?9 and hd(acc) == "-") ->
+          # https://drafts.csswg.org/cssom/#escape-a-character-as-code-point
+          ["\\", Integer.to_string(char, 16), " "]
+
+        # If the character is not handled by one of the above rules and is
+        # greater than or equal to U+0080, is `-` (U+002D) or `_` (U+005F), or
+        # is in one of the ranges [0-9] (U+0030 to U+0039), [A-Z] (U+0041 to
+        # U+005A), or [a-z] (U+0061 to U+007A), […]
+        char >= 0x0080 or char in [?-, ?_] or char in ?0..?9 or char in ?A..?Z or char in ?a..?z ->
+          # the character itself
+          <<char::utf8>>
+
+        true ->
+          # Otherwise, the escaped character.
+          # https://drafts.csswg.org/cssom/#escape-a-character
+          ["\\", <<char::utf8>>]
+      end
+
+    do_escape_chars(rest, index + 1, [escaped | acc])
+  end
+end
diff --git a/test/floki/css_escape_test.exs b/test/floki/css_escape_test.exs
new file mode 100644
index 00000000..f17f3003
--- /dev/null
+++ b/test/floki/css_escape_test.exs
@@ -0,0 +1,79 @@
+defmodule Floki.CSSEscapeTest do
+  use ExUnit.Case, async: true
+
+  doctest Floki.CSSEscape
+
+  test "null character" do
+    assert Floki.CSSEscape.escape(<<0>>) == <<0xFFFD::utf8>>
+    assert Floki.CSSEscape.escape("a\u0000") == "a\ufffd"
+    assert Floki.CSSEscape.escape("\u0000b") == "\ufffdb"
+    assert Floki.CSSEscape.escape("a\u0000b") == "a\ufffdb"
+  end
+
+  test "replacement character" do
+    assert Floki.CSSEscape.escape(<<0xFFFD::utf8>>) == <<0xFFFD::utf8>>
+    assert Floki.CSSEscape.escape("a\ufffd") == "a\ufffd"
+    assert Floki.CSSEscape.escape("\ufffdb") == "\ufffdb"
+    assert Floki.CSSEscape.escape("a\ufffdb") == "a\ufffdb"
+  end
+
+  test "invalid input" do
+    assert_raise ArgumentError, fn -> Floki.CSSEscape.escape(nil) end
+  end
+
+  test "control characters" do
+    assert Floki.CSSEscape.escape(<<0x01, 0x02, 0x1E, 0x1F>>) == "\\1 \\2 \\1E \\1F "
+  end
+
+  test "leading digit" do
+    for {digit, expected} <- Enum.zip(0..9, ~w(30 31 32 33 34 35 36 37 38 39)) do
+      assert Floki.CSSEscape.escape("#{digit}a") == "\\#{expected} a"
+    end
+  end
+
+  test "non-leading digit" do
+    for digit <- 0..9 do
+      assert Floki.CSSEscape.escape("a#{digit}b") == "a#{digit}b"
+    end
+  end
+
+  test "leading hyphen and digit" do
+    for {digit, expected} <- Enum.zip(0..9, ~w(30 31 32 33 34 35 36 37 38 39)) do
+      assert Floki.CSSEscape.escape("-#{digit}a") == "-\\#{expected} a"
+    end
+  end
+
+  test "hyphens" do
+    assert Floki.CSSEscape.escape("-") == "\\-"
+    assert Floki.CSSEscape.escape("-a") == "-a"
+    assert Floki.CSSEscape.escape("--") == "--"
+    assert Floki.CSSEscape.escape("--a") == "--a"
+  end
+
+  test "non-ASCII and special characters" do
+    assert Floki.CSSEscape.escape("🤷🏻‍♂️-_©") == "🤷🏻‍♂️-_©"
+
+    assert Floki.CSSEscape.escape(
+             <<0x7F,
+               "\u0080\u0081\u0082\u0083\u0084\u0085\u0086\u0087\u0088\u0089\u008a\u008b\u008c\u008d\u008e\u008f\u0090\u0091\u0092\u0093\u0094\u0095\u0096\u0097\u0098\u0099\u009a\u009b\u009c\u009d\u009e\u009f">>
+           ) ==
+             "\\7F \u0080\u0081\u0082\u0083\u0084\u0085\u0086\u0087\u0088\u0089\u008a\u008b\u008c\u008d\u008e\u008f\u0090\u0091\u0092\u0093\u0094\u0095\u0096\u0097\u0098\u0099\u009a\u009b\u009c\u009d\u009e\u009f"
+
+    assert Floki.CSSEscape.escape("\u00a0\u00a1\u00a2") == "\u00a0\u00a1\u00a2"
+  end
+
+  test "alphanumeric characters" do
+    assert Floki.CSSEscape.escape("a0123456789b") == "a0123456789b"
+    assert Floki.CSSEscape.escape("abcdefghijklmnopqrstuvwxyz") == "abcdefghijklmnopqrstuvwxyz"
+    assert Floki.CSSEscape.escape("ABCDEFGHIJKLMNOPQRSTUVWXYZ") == "ABCDEFGHIJKLMNOPQRSTUVWXYZ"
+  end
+
+  test "space and exclamation mark" do
+    assert Floki.CSSEscape.escape(<<0x20, 0x21, 0x78, 0x79>>) == "\\ \\!xy"
+  end
+
+  test "unicode characters" do
+    # astral symbol (U+1D306 TETRAGRAM FOR CENTRE)
+    assert Floki.CSSEscape.escape(<<0x1D306::utf8>>) == <<0x1D306::utf8>>
+  end
+end