diff --git a/book/src/generated/static-cmd.md b/book/src/generated/static-cmd.md
index af7515b8e..532c032cb 100644
--- a/book/src/generated/static-cmd.md
+++ b/book/src/generated/static-cmd.md
@@ -45,6 +45,10 @@
 | `extend_parent_node_start` | Extend to beginning of the parent node | select: `` <A-b> `` |
 | `find_till_char` | Move till next occurrence of char | normal: `` t `` |
 | `find_next_char` | Move to next occurrence of char | normal: `` f `` |
+| `find_next_pair` | Move to next occurrence of 2 chars | normal: `` L `` |
+| `find_prev_pair` | Move to next occurrence of 2 chars | normal: `` H `` |
+| `extend_next_pair` | Extend to next occurrence of 2 chars | select: `` L `` |
+| `extend_prev_pair` | Extend to next occurrence of 2 chars | select: `` H `` |
 | `extend_till_char` | Extend till next occurrence of char | select: `` t `` |
 | `extend_next_char` | Extend to next occurrence of char | select: `` f `` |
 | `till_prev_char` | Move till previous occurrence of char | normal: `` T `` |
diff --git a/book/src/keymap.md b/book/src/keymap.md
index 2797eaee2..7008d625e 100644
--- a/book/src/keymap.md
+++ b/book/src/keymap.md
@@ -49,8 +49,10 @@ Normal mode is the default mode when you launch helix. You can return to it from
 | `E`                   | Move next WORD end                                 | `move_next_long_word_end`   |
 | `t`                   | Find 'till next char                               | `find_till_char`            |
 | `f`                   | Find next char                                     | `find_next_char`            |
+| `L`                   | Find next 2 chars                                  | `find_next_pair`            |
 | `T`                   | Find 'till previous char                           | `till_prev_char`            |
 | `F`                   | Find previous char                                 | `find_prev_char`            |
+| `H`                   | Find prev 2 chars                                  | `find_prev_pair`            |
 | `G`                   | Go to line number `<n>`                            | `goto_line`                 |
 | `Alt-.`               | Repeat last motion (`f`, `t`, `m`, `[` or `]`)     | `repeat_last_motion`        |
 | `Home`                | Move to the start of the line                      | `goto_line_start`           |
diff --git a/helix-core/src/search.rs b/helix-core/src/search.rs
index 81cb41293..ddab8aed9 100644
--- a/helix-core/src/search.rs
+++ b/helix-core/src/search.rs
@@ -1,4 +1,4 @@
-use crate::RopeSlice;
+use crate::{line_ending::line_end_char_index, movement::Direction, RopeSlice};
 
 // TODO: switch to std::str::Pattern when it is stable.
 pub trait CharMatcher {
@@ -65,3 +65,179 @@ pub fn find_nth_prev(text: RopeSlice, ch: char, mut pos: usize, n: usize) -> Opt
 
     Some(pos)
 }
+
+#[derive(Copy, Clone)]
+pub enum PairMatcher<'a> {
+    Char(char),
+    LineEnding(&'a str),
+}
+
+pub fn find_nth_pair(
+    text: RopeSlice,
+    pair_matcher_left: PairMatcher,
+    pair_matcher_right: PairMatcher,
+    pos: usize,
+    n: usize,
+    direction: Direction,
+) -> Option<usize> {
+    if pos >= text.len_chars() || n == 0 {
+        return None;
+    }
+
+    let is_forward = direction == Direction::Forward;
+    let direction_multiplier = if is_forward { 1 } else { -1 };
+
+    match (pair_matcher_left, pair_matcher_right) {
+        (PairMatcher::Char(ch_left), PairMatcher::Char(ch_right)) => {
+            let chars = text.chars_at(pos);
+
+            let mut chars = if is_forward {
+                chars.peekable()
+            } else {
+                chars.reversed().peekable()
+            };
+
+            let mut offset = 0;
+
+            for _ in 0..n {
+                loop {
+                    let ch_next = chars.next()?;
+                    let ch_peek = chars.peek()?;
+
+                    offset += 1;
+
+                    let matches_char = if is_forward {
+                        ch_left == ch_next && ch_right == *ch_peek
+                    } else {
+                        ch_right == ch_next && ch_left == *ch_peek
+                    };
+
+                    if matches_char {
+                        break;
+                    }
+                }
+            }
+
+            let offs = offset * direction_multiplier;
+            let new_pos: usize = (pos as isize + offs)
+                .try_into()
+                .expect("Character offset cannot exceed character count");
+
+            Some(new_pos - 1)
+        }
+        (PairMatcher::Char(ch_left), PairMatcher::LineEnding(eol)) => {
+            let start_line = text.char_to_line(pos);
+            let start_line = if pos >= line_end_char_index(&text, start_line) {
+                // if our cursor is currently on a character just before the eol, or on the eol
+                // we start searching from the next line, instead of from the current line.
+                start_line + eol.len()
+            } else {
+                start_line
+            };
+
+            let mut lines = if is_forward {
+                text.lines_at(start_line).enumerate()
+            } else {
+                text.lines_at(start_line).reversed().enumerate()
+            };
+
+            if !is_forward {
+                // skip the line we are currently on when going backward
+                lines.next();
+            }
+
+            let mut matched_count = 0;
+            for (traversed_lines, _line) in lines {
+                let current_line = (start_line as isize
+                    + (traversed_lines as isize * direction_multiplier))
+                    as usize;
+
+                let ch_opposite_eol_i = if is_forward {
+                    line_end_char_index(&text, current_line).saturating_sub(eol.len())
+                } else {
+                    text.line_to_char(current_line)
+                };
+
+                let ch_opposite_eol = text.char(ch_opposite_eol_i);
+
+                if ch_opposite_eol == ch_left {
+                    matched_count += 1;
+                    if matched_count == n {
+                        return Some(ch_opposite_eol_i - if is_forward { 0 } else { 1 });
+                    }
+                }
+            }
+
+            None
+        }
+        (PairMatcher::LineEnding(eol), PairMatcher::Char(ch_right)) => {
+            // Search starting from the beginning of the next or previous line
+            let start_line = text.char_to_line(pos) + (is_forward as usize);
+
+            let lines = if is_forward {
+                text.lines_at(start_line).enumerate()
+            } else {
+                text.lines_at(start_line).reversed().enumerate()
+            };
+
+            let mut matched_count = 0;
+            for (traversed_lines, _line) in lines {
+                let current_line = (start_line as isize
+                    + (traversed_lines as isize * direction_multiplier))
+                    as usize;
+
+                let ch_opposite_eol_i = if is_forward {
+                    // eol, THEN character at the beginning of the current line
+                    text.line_to_char(current_line)
+                } else {
+                    // character at the end of the previous line, THEN eol
+                    line_end_char_index(&text, current_line - 1) - eol.len()
+                };
+
+                let ch_opposite_eol = text.get_char(ch_opposite_eol_i)?;
+
+                if ch_opposite_eol == ch_right {
+                    matched_count += 1;
+                    if matched_count == n {
+                        return Some(ch_opposite_eol_i - (is_forward as usize));
+                    }
+                }
+            }
+
+            None
+        }
+        (PairMatcher::LineEnding(eol), PairMatcher::LineEnding(_)) => {
+            // Search starting from the beginning of the
+            // line after the current one
+            let start_line = text.char_to_line(pos) + 1;
+
+            let mut lines = if is_forward {
+                text.lines_at(start_line).enumerate()
+            } else {
+                text.lines_at(start_line).reversed().enumerate()
+            };
+
+            if !is_forward {
+                // skip the line we are currently on when going backward
+                lines.next();
+            }
+
+            let mut matched_count = 0;
+            for (traversed_lines, _line) in lines {
+                let current_line = (start_line as isize
+                    + (traversed_lines as isize * direction_multiplier))
+                    as usize;
+                let current_line = text.line_to_char(current_line);
+                let current_line_end = current_line + eol.len();
+                if text.slice(current_line..current_line_end).as_str()? == eol {
+                    matched_count += 1;
+                    if matched_count == n {
+                        return Some(current_line - eol.len());
+                    }
+                }
+            }
+
+            None
+        }
+    }
+}
diff --git a/helix-core/src/selection.rs b/helix-core/src/selection.rs
index 1db2d619e..8d16c7095 100644
--- a/helix-core/src/selection.rs
+++ b/helix-core/src/selection.rs
@@ -329,6 +329,18 @@ impl Range {
     //--------------------------------
     // Block-cursor methods.
 
+    /// Gets the left-side position of the block cursor.
+    /// Not grapheme-aware. For the grapheme-aware version, use [`Range::cursor`]
+    #[must_use]
+    #[inline]
+    pub fn char_cursor(self) -> usize {
+        if self.head > self.anchor {
+            self.head - 1
+        } else {
+            self.head
+        }
+    }
+
     /// Gets the left-side position of the block cursor.
     #[must_use]
     #[inline]
diff --git a/helix-term/src/commands.rs b/helix-term/src/commands.rs
index 2e15dcdcc..5ecaaad51 100644
--- a/helix-term/src/commands.rs
+++ b/helix-term/src/commands.rs
@@ -343,6 +343,10 @@ impl MappableCommand {
         extend_parent_node_start, "Extend to beginning of the parent node",
         find_till_char, "Move till next occurrence of char",
         find_next_char, "Move to next occurrence of char",
+        find_next_pair, "Move to next occurrence of 2 chars",
+        find_prev_pair, "Move to next occurrence of 2 chars",
+        extend_next_pair, "Extend to next occurrence of 2 chars",
+        extend_prev_pair, "Extend to next occurrence of 2 chars",
         extend_till_char, "Extend till next occurrence of char",
         extend_next_char, "Extend to next occurrence of char",
         till_prev_char, "Move till previous occurrence of char",
@@ -1553,7 +1557,83 @@ fn find_char(cx: &mut Context, direction: Direction, inclusive: bool, extend: bo
     })
 }
 
-//
+fn find_char_pair(cx: &mut Context, direction: Direction, extend: bool) {
+    // TODO: count is reset to 1 before next key so we move it into the closure here.
+    // Would be nice to carry over.
+    let count = cx.count();
+    let eof = doc!(cx.editor).line_ending.as_str();
+
+    // need to wait for next key
+    // TODO: should this be done by grapheme rather than char?  For example,
+    // we can't properly handle the line-ending CRLF case here in terms of char.
+    cx.on_next_key(move |cx, event| {
+        let ch = match event {
+            KeyEvent {
+                code: KeyCode::Enter,
+                ..
+            } => search::PairMatcher::LineEnding(eof),
+
+            KeyEvent {
+                code: KeyCode::Tab, ..
+            } => search::PairMatcher::Char('\t'),
+
+            KeyEvent {
+                code: KeyCode::Char(ch),
+                ..
+            } => search::PairMatcher::Char(ch),
+            _ => return,
+        };
+
+        cx.on_next_key(move |cx, event| {
+            let ch_2 = match event {
+                KeyEvent {
+                    code: KeyCode::Enter,
+                    ..
+                } => search::PairMatcher::LineEnding(eof),
+
+                KeyEvent {
+                    code: KeyCode::Tab, ..
+                } => search::PairMatcher::Char('\t'),
+
+                KeyEvent {
+                    code: KeyCode::Char(ch),
+                    ..
+                } => search::PairMatcher::Char(ch),
+                _ => return,
+            };
+            let motion = move |editor: &mut Editor| {
+                let (view, doc) = current!(editor);
+                let text = doc.text().slice(..);
+                let selection = doc.selection(view.id).clone();
+                let selection = match direction {
+                    Direction::Forward => selection.transform(|range| {
+                        search::find_nth_pair(text, ch, ch_2, range.char_cursor(), count, direction)
+                            .map_or(range, |pos| {
+                                if extend {
+                                    Range::new(range.from(), pos + 2)
+                                } else {
+                                    Range::new(pos, pos + 2)
+                                }
+                            })
+                    }),
+                    Direction::Backward => selection.transform(|range| {
+                        search::find_nth_pair(text, ch, ch_2, range.char_cursor(), count, direction)
+                            .map_or(range, |pos| {
+                                if extend {
+                                    Range::new(pos + 2, range.to())
+                                } else {
+                                    Range::new(pos + 2, pos)
+                                }
+                            })
+                    }),
+                };
+                doc.set_selection(view.id, selection);
+            };
+
+            cx.editor.apply_motion(motion);
+        })
+    })
+}
 
 #[inline]
 fn find_char_impl<F, M: CharMatcher + Clone + Copy>(
@@ -1566,20 +1646,12 @@ fn find_char_impl<F, M: CharMatcher + Clone + Copy>(
 ) where
     F: Fn(RopeSlice, M, usize, usize, bool) -> Option<usize> + 'static,
 {
+    // TODO: make this grapheme-aware
     let (view, doc) = current!(editor);
     let text = doc.text().slice(..);
 
     let selection = doc.selection(view.id).clone().transform(|range| {
-        // TODO: use `Range::cursor()` here instead.  However, that works in terms of
-        // graphemes, whereas this function doesn't yet.  So we're doing the same logic
-        // here, but just in terms of chars instead.
-        let search_start_pos = if range.anchor < range.head {
-            range.head - 1
-        } else {
-            range.head
-        };
-
-        search_fn(text, char_matcher, search_start_pos, count, inclusive).map_or(range, |pos| {
+        search_fn(text, char_matcher, range.char_cursor(), count, inclusive).map_or(range, |pos| {
             if extend {
                 range.put_cursor(text, pos, true)
             } else {
@@ -1627,6 +1699,22 @@ fn find_prev_char_impl(
     }
 }
 
+fn find_next_pair(cx: &mut Context) {
+    find_char_pair(cx, Direction::Forward, false)
+}
+
+fn find_prev_pair(cx: &mut Context) {
+    find_char_pair(cx, Direction::Backward, false)
+}
+
+fn extend_next_pair(cx: &mut Context) {
+    find_char_pair(cx, Direction::Forward, true)
+}
+
+fn extend_prev_pair(cx: &mut Context) {
+    find_char_pair(cx, Direction::Backward, true)
+}
+
 fn find_till_char(cx: &mut Context) {
     find_char(cx, Direction::Forward, false, false);
 }
diff --git a/helix-term/src/keymap/default.rs b/helix-term/src/keymap/default.rs
index e160b2246..fe64eea21 100644
--- a/helix-term/src/keymap/default.rs
+++ b/helix-term/src/keymap/default.rs
@@ -11,6 +11,9 @@ pub fn default() -> HashMap<Mode, KeyTrie> {
         "k" | "up" => move_visual_line_up,
         "l" | "right" => move_char_right,
 
+        "L" => find_next_pair,
+        "H" => find_prev_pair,
+
         "t" => find_till_char,
         "f" => find_next_char,
         "T" => till_prev_char,
@@ -343,6 +346,9 @@ pub fn default() -> HashMap<Mode, KeyTrie> {
         "k" | "up" => extend_visual_line_up,
         "l" | "right" => extend_char_right,
 
+        "L" => extend_next_pair,
+        "H" => extend_prev_pair,
+
         "w" => extend_next_word_start,
         "b" => extend_prev_word_start,
         "e" => extend_next_word_end,
diff --git a/helix-term/tests/test/commands.rs b/helix-term/tests/test/commands.rs
index 2af1a054f..c09581526 100644
--- a/helix-term/tests/test/commands.rs
+++ b/helix-term/tests/test/commands.rs
@@ -6,6 +6,157 @@ mod insert;
 mod movement;
 mod write;
 
+#[tokio::test(flavor = "multi_thread")]
+async fn find_prev_pair() -> anyhow::Result<()> {
+    // finds prev pair of 2 letters
+    test((
+        indoc! {"\
+            hi
+            hi
+            #[hi|]#
+            hi"},
+        "Hhi",
+        indoc! {"\
+            hi
+            #[|hi]#
+            hi
+            hi"},
+    ))
+    .await?;
+    // finds 3rd prev pair of 2 letters
+    test((
+        indoc! {"\
+            hi
+            hi
+            hi
+            #[hi|]#"},
+        "3Hhi",
+        indoc! {"\
+            #[|hi]#
+            hi
+            hi
+            hi"},
+    ))
+    .await?;
+    // finds prev two newlines
+    test((
+        indoc! {"\
+            hi
+            hi
+            hi
+
+            #[hi|]#"},
+        "H<ret><ret>",
+        indoc! {"\
+            hi
+            hi
+            hi#[|
+
+            ]#hi"},
+    ))
+    .await?;
+
+    Ok(())
+}
+
+#[tokio::test(flavor = "multi_thread")]
+async fn find_next_pair() -> anyhow::Result<()> {
+    // finds next pair of 2 letters (non-extend)
+    test((
+        indoc! {"\
+            #[hi|]#
+            hi
+            hi
+            hi"},
+        "Lhi",
+        indoc! {"\
+            hi
+            #[hi|]#
+            hi
+            hi"},
+    ))
+    .await?;
+    // finds next pair of 2 letters (extend)
+    test((
+        indoc! {"\
+            hi
+            #[hi|]#
+            hi
+            hi"},
+        "vLhi",
+        indoc! {"\
+            hi
+            #[hi
+            hi|]#
+            hi"},
+    ))
+    .await?;
+    // finds 3rd next pair of 2 letters (non-extend)
+    test((
+        indoc! {"\
+            #[hi|]#
+            hi
+            hi
+            hi"},
+        "3Lhi",
+        indoc! {"\
+            hi
+            hi
+            hi
+            #[hi|]#"},
+    ))
+    .await?;
+    // finds 3rd next pair of 2 letters (extend)
+    test((
+        indoc! {"\
+            #[hi|]#
+            hi
+            hi
+            hi"},
+        "v3Lhi",
+        indoc! {"\
+            #[hi
+            hi
+            hi
+            hi|]#"},
+    ))
+    .await?;
+    // finds next char followed by newline
+    test((
+        indoc! {"\
+            #[hi|]#
+            hi
+            hi
+            hi"},
+        "L<ret>h",
+        indoc! {"\
+            hi#[
+            h|]#i
+            hi
+            hi"},
+    ))
+    .await?;
+    // finds next char followed by newline
+    test((
+        indoc! {"\
+            #[hi|]#
+            hi
+            hi
+
+            hi"},
+        "L<ret><ret>",
+        indoc! {"\
+            hi
+            hi
+            hi#[
+
+            |]#hi"},
+    ))
+    .await?;
+
+    Ok(())
+}
+
 #[tokio::test(flavor = "multi_thread")]
 async fn search_selection_detect_word_boundaries_at_eof() -> anyhow::Result<()> {
     // <https://github.com/helix-editor/helix/issues/12609>