diff --git a/lib/experimental/html/class-wp-html-processor-scan-state.php b/lib/experimental/html/class-wp-html-processor-scan-state.php
new file mode 100644
index 00000000000000..5f02ec86245c54
--- /dev/null
+++ b/lib/experimental/html/class-wp-html-processor-scan-state.php
@@ -0,0 +1,50 @@
+<?php
+/**
+ * HTML Processor Scan State: Track opening tags and scanning depth.
+ *
+ * @package WordPress
+ * @subpackage HTML
+ * @since 6.2.0
+ */
+
+/**
+ * Track opening tags and scanning depth.
+ *
+ * This class is for internal usage of the WP_HTML_Processor class.
+ *
+ * @access private
+ * @since 6.2.0
+ *
+ * @see WP_HTML_Processor
+ */
+class WP_HTML_Processor_Scan_State {
+	/**
+	 * The maximum number of tags we'll traverse in search of a matching closing tag.
+	 *
+	 * @var integer
+	 */
+	public $budget = 1000;
+
+	/**
+	 * A stack of the opening tags that we have visited.
+	 *
+	 * @var string[]
+	 */
+	public $open_tags = array();
+
+	/**
+	 * The depth of nested tags at which we expect to find the matching closing tag.
+	 *
+	 * @var int
+	 */
+	public $match_depth = null;
+
+	/**
+	 * The depth of nested opening tags, counted from where we started.
+	 *
+	 * @return int The depth of nested tags.
+	 */
+	public function relative_depth() {
+		return count( $this->open_tags );
+	}
+}
diff --git a/lib/experimental/html/class-wp-html-processor.php b/lib/experimental/html/class-wp-html-processor.php
new file mode 100644
index 00000000000000..482065d813623c
--- /dev/null
+++ b/lib/experimental/html/class-wp-html-processor.php
@@ -0,0 +1,292 @@
+<?php
+/**
+ * Scans through an HTML document to find specific tags, then
+ * transforms those tags by adding, removing, or updating the
+ * values of the HTML attributes within that tag (opener).
+ * Furthermore finds the matching closing tag for a given
+ * opening tag, and retrieves the content in between them.
+ *
+ * @TODO: Handle self-closing foreign elements.
+ * @TODO: Detect non-normative HTML input.
+ * @TODO: Consider parsing non-normative HTML input, support adoption agency algorithm.
+ * @TODO: Figure out how multiple external states can conflict.
+ *
+ * If we support non-normative HTML we can probably handle significantly more
+ * HTML without introducing unexpected results, but I'm not sure yet if we can
+ * handle HTML the same way as the browser, because the section in HTML5 spec
+ * dealing with errors is itself "non-normative" and only issues a few examles.
+ *
+ * Not yet clear is if browsers are full of special one-off cases for "invalid"
+ * input. E.g. it's clear to me how to handle `</b></b></b>` but not clear to how
+ * handle `</p></p></p>` given that `<b>` is a formatting element but `<p>` is
+ * not, that `<p>` itself is a special element.
+ *
+ * @package WordPress
+ * @subpackage HTML
+ * @since 6.2.0
+ */
+
+/**
+ * Processes an input HTML document by applying a specified set of patches
+ * to that input. Retrieves content between matching opening and closing tags.
+ * Tokenizes HTML but does not fully parse the input document.
+ *
+ * ## Usage
+ *
+ * Note that this is a subclass of `WP_HTML_Tag_Processor`. Most of the
+ * functionality of this class is thus covered by `WP_HTML_Tag_Processor`'s
+ * documentation.
+ * The following documentation covers the additional features added by
+ * `WP_HTML_Processor`.
+ *
+ * ### Retrieving content
+ *
+ * When on an opening tag, it's possible to retrieve the content enclosed between
+ * that opening tag and its matching closing tag.
+ *
+ * Example:
+ * ```php
+ *     $html = '<div id="outer"><div>Inner div content</div><img></div>';
+ *     $tags = new WP_HTML_Processor( $html );
+ *     $tags->next_tag( [ 'tag_name' => 'div' ];
+ *     $label = $tags->get_content_inside_balanced_tags();
+ *     // $label === '<div>Inner div content</div><img>'
+ *     }
+ * ```
+ *
+ * @see WP_HTML_Tag_Processor
+ */
+class WP_HTML_Processor extends WP_HTML_Tag_Processor {
+	/**
+	 * Create a new tracking state for, based on the current opening tag.
+	 *
+	 * @return WP_HTML_Processor_Scan_State
+	 */
+	public function new_state() {
+		$state    = new WP_HTML_Processor_Scan_State();
+		$tag_name = $this->get_tag();
+
+		if ( ! self::is_html_void_element( $tag_name ) && ! $this->is_tag_closer() ) {
+			$state->open_tags[] = $tag_name;
+		}
+
+		return $state;
+	}
+
+	/**
+	 * Find the matching closing tag for an opening tag.
+	 *
+	 * When called while on an open tag, traverse the HTML until we find
+	 * the matching closing tag, respecting any in-between content, including
+	 * nested tags of the same name. Return false when called on a closing or
+	 * void tag, or if no matching closing tag was found.
+	 *
+	 * @param WP_HTML_Processor_Scan_State $state Tracking state.
+	 * @param array|string                 $query Query criteria for the closing tag.
+	 * @return bool                        True if a matching closing tag was found.
+	 *
+	 * @see WP_HTML_Tag_Processor::parse_query
+	 */
+	public function balanced_next( WP_HTML_Processor_Scan_State $state, $query = null ) {
+		while ( $this->next_tag( array( 'tag_closers' => 'visit' ) ) && $state->budget-- > 0 ) {
+			$tag_name  = $this->get_tag();
+			$is_closer = $this->is_tag_closer();
+			$is_void   = self::is_html_void_element( $tag_name );
+			$type      = self::classify_tag_type( $is_closer, $is_void );
+
+			/*
+			 * Step 1. Update the stack of open tags.
+			 *
+			 * If and when we add more complete HTML parsing support we will also
+			 * need to track the stack of active formats so that we can properly
+			 * handle missing tags and overlapping tags.
+			 */
+
+			switch ( $type ) {
+				case 'void':
+					/*
+					 * Void tags (such as <img>) can't have children and so we
+					 * won't push or pop them from the stack of open tags.
+					 *
+					 * If and when we support self-closing foreign tags we would
+					 * need to separately track those, but their behavior matches
+					 * this case. The self-closing flag is ignored for HTML5 tags.
+					 */
+					if ( 0 === $state->relative_depth() ) {
+						return false;
+					}
+
+					break;
+
+				case 'opener':
+					$state->open_tags[] = $tag_name;
+					break;
+
+				case 'closer':
+					$last_tag = array_pop( $state->open_tags );
+
+					/*
+					 * Currently we can only support fully-normative and balanced HTML5.
+					 * If we encounter anything we don't expect then we will bail. In a
+					 * future update we may perform more careful HTML parsing and unlock
+					 * navigating through non-normative documents.
+					 */
+					if ( $last_tag !== $tag_name ) {
+						return false;
+					}
+
+					/*
+					 * Step 2. Bail if we've reached the end of the tag in which we started.
+					 */
+					if ( 0 === $state->relative_depth() ) {
+						return false;
+					}
+
+					break;
+			}
+
+			/*
+			 * Void elements don't enter the stack, but they do exist in the
+			 * depth hierarchy, so we have to temporarily account for that.
+			 *
+			 * We could have followed the approach in the HTML5 spec by appending
+			 * the void tag to the stack of open tags, and then remember to pop it
+			 * when existing this function, but by tracking it like this we don't
+			 * have to remember to do that.
+			 */
+			$depth = 'void' === $type
+				? $state->relative_depth() + 1
+				: $state->relative_depth();
+
+			/*
+			 * Step 3. Determine if we have a matching tag. In addition to the query
+			 * we pass along to the underlying tag processor we're going to allow
+			 * specifying the relative depth for a match. For example, a CSS child
+			 * combinator would specify that a match must have a relative depth of 1,
+			 * indicating that it's a direct child of the surrounding element, whereas
+			 * the descendant selector could match at any depth and so sets this to `null`.
+			 * To prevent matching _above_ a tag we rely on the `bail_depth` to stop
+			 * searching once we've exited the tag on which we started, or reach its parent.
+			 */
+
+			if ( ! isset( $state->match_depth ) || $state->match_depth + 1 === $depth ) {
+				$this->parse_query( $query );
+				if ( $this->matches() ) {
+					return true;
+				}
+			}
+		}
+
+		return false;
+	}
+
+	/**
+	 * Return the content between two balanced tags.
+	 *
+	 * When called on an opening tag, return the HTML content found between
+	 * that opening tag and its matching closing tag.
+	 *
+	 * @return string The content between the current opening and its matching closing tag.
+	 */
+	public function get_content_inside_balanced_tags() {
+		static $start_name = null;
+		static $end_name   = null;
+
+		if ( null === $start_name || array_key_exists( $start_name, $this->bookmarks ) ) {
+			$rand_id    = rand( 1, PHP_INT_MAX );
+			$start_name = "start_{$rand_id}";
+		}
+
+		if ( null === $end_name || array_key_exists( $end_name, $this->bookmarks ) ) {
+			$rand_id  = rand( 1, PHP_INT_MAX );
+			$end_name = "start_{$rand_id}";
+		}
+
+		$this->set_bookmark( $start_name );
+
+		$state = self::new_state();
+		while ( $this->balanced_next( $state ) ) {
+			continue;
+		}
+
+		$this->set_bookmark( $end_name );
+		$content = $this->content_inside_bookmarks( $start_name, $end_name );
+		$this->seek( $start_name );
+
+		$this->release_bookmark( $start_name );
+		$this->release_bookmark( $end_name );
+
+		return $content;
+	}
+
+	/**
+	 * Return the content between two bookmarks.
+	 *
+	 * @param WP_HTML_Span $start_bookmark The bookmark marking the start of the content.
+	 * @param WP_HTML_Span $end_bookmark   The bookmark marking the start of the content.
+	 * @return string|null                 The content between the two bookmarks.
+	 *                                     Null if either of the bookmarks isn't set.
+	 */
+	private function content_inside_bookmarks( $start_bookmark, $end_bookmark ) {
+		if ( ! isset( $this->bookmarks[ $start_bookmark ], $this->bookmarks[ $end_bookmark ] ) ) {
+			return null;
+		}
+
+		$start = $this->bookmarks[ $start_bookmark ];
+		$end   = $this->bookmarks[ $end_bookmark ];
+
+		return substr( $this->get_updated_html(), $start->end + 1, $end->start - $start->end - 2 );
+	}
+
+	/*
+	 * HTML-related Utility Functions
+	 */
+
+	/**
+	 * Classify a given HTML tag type.
+	 *
+	 * Return 'opener' for an opening element, 'closer' for a closing element,
+	 * and 'void' for a void element.
+	 *
+	 * @param bool $is_closer Whether the current element is a closing element.
+	 * @param bool $is_void Whether the current element is a void element.
+	 * @return 'opener'|'closer'|'void' The type of element in question.
+	 */
+	public static function classify_tag_type( $is_closer, $is_void ) {
+		if ( $is_void ) {
+			return 'void';
+		}
+
+		return $is_closer ? 'closer' : 'opener';
+	}
+
+	/**
+	 * Whether a given HTML element is void (e.g. <br>).
+	 *
+	 * @param string $tag_name The element in question.
+	 * @return bool True if the element is void.
+	 *
+	 * @see https://html.spec.whatwg.org/#elements-2
+	 */
+	public static function is_html_void_element( $tag_name ) {
+		switch ( $tag_name ) {
+			case 'AREA':
+			case 'BASE':
+			case 'BR':
+			case 'COL':
+			case 'EMBED':
+			case 'HR':
+			case 'IMG':
+			case 'INPUT':
+			case 'LINK':
+			case 'META':
+			case 'SOURCE':
+			case 'TRACK':
+			case 'WBR':
+				return true;
+
+			default:
+				return false;
+		}
+	}
+}
diff --git a/lib/experimental/html/class-wp-html-tag-processor.php b/lib/experimental/html/class-wp-html-tag-processor.php
index 433c16a150806c..8bb4e52fde0685 100644
--- a/lib/experimental/html/class-wp-html-tag-processor.php
+++ b/lib/experimental/html/class-wp-html-tag-processor.php
@@ -1900,7 +1900,7 @@ public function get_updated_html() {
 	 *     @type string      $tag_closers  "visit" or "skip": whether to stop on tag closers, e.g. </div>.
 	 * }
 	 */
-	private function parse_query( $query ) {
+	protected function parse_query( $query ) {
 		if ( null !== $query && $query === $this->last_query ) {
 			return;
 		}
@@ -1947,7 +1947,7 @@ private function parse_query( $query ) {
 	 *
 	 * @return boolean
 	 */
-	private function matches() {
+	protected function matches() {
 		if ( $this->is_closing_tag && ! $this->stop_on_tag_closers ) {
 			return false;
 		}
diff --git a/lib/experimental/html/wp-html.php b/lib/experimental/html/wp-html.php
index dd3aeb7af45ae9..68e7c57c7aaf3e 100644
--- a/lib/experimental/html/wp-html.php
+++ b/lib/experimental/html/wp-html.php
@@ -21,3 +21,11 @@
 if ( ! class_exists( 'WP_HTML_Tag_Processor' ) ) {
 	require_once __DIR__ . '/class-wp-html-tag-processor.php';
 }
+
+if ( ! class_exists( 'WP_HTML_Processor_Scan_State' ) ) {
+	require_once __DIR__ . '/class-wp-html-processor-scan-state.php';
+}
+
+if ( ! class_exists( 'WP_HTML_Processor' ) ) {
+	require_once __DIR__ . '/class-wp-html-processor.php';
+}
diff --git a/phpunit/html/wp-html-processor-test.php b/phpunit/html/wp-html-processor-test.php
new file mode 100644
index 00000000000000..5582a988bf067e
--- /dev/null
+++ b/phpunit/html/wp-html-processor-test.php
@@ -0,0 +1,229 @@
+<?php
+
+/**
+ * Unit tests covering WP_HTML_Processor functionality.
+ *
+ * @package WordPress
+ * @subpackage HTML
+ */
+
+require_once __DIR__ . '/../../lib/experimental/html/wp-html.php';
+
+if ( ! function_exists( 'esc_attr' ) ) {
+	function esc_attr( $s ) {
+		return htmlentities( $s, ENT_QUOTES, null, false ); }
+}
+
+if ( ! class_exists( 'WP_UnitTestCase' ) ) {
+	class WP_UnitTestCase extends PHPUnit\Framework\TestCase {}
+}
+
+/**
+ * @group html-processor
+ *
+ * @coversDefaultClass WP_HTML_Processor
+ */
+class WP_HTML_Processor_Test extends WP_UnitTestCase {
+	public function test_find_descendant_tag() {
+		$tags = new WP_HTML_Processor( '<div>outside</div><section><div><img>inside</div></section>' );
+
+		$tags->next_tag( 'div' );
+		$state = $tags->new_state();
+		$this->assertFalse( $tags->balanced_next( $state, 'img' ) );
+
+		$this->assertTrue( $tags->next_tag( 'div' ) );
+		$state = $tags->new_state();
+		$this->assertTrue( $tags->balanced_next( $state, 'img' ) );
+	}
+
+	public function test_find_immediate_child_tag() {
+		$tags = new WP_HTML_Processor( '<div><div><div><img></div></div></div>' );
+
+		$tags->next_tag( 'div' );
+		$state              = $tags->new_state();
+		$state->match_depth = 1;
+		$this->assertFalse( $tags->balanced_next( $state, 'img' ) );
+	}
+
+	public function test_find_immediate_child_tag2() {
+		$tags = new WP_HTML_Processor( '<div><div><div><img></div></div><img wanted></div>' );
+
+		$tags->next_tag( 'div' );
+		$state              = $tags->new_state();
+		$state->match_depth = 1;
+		$this->assertTrue( $tags->balanced_next( $state, 'img' ), 'Did not find the wanted <img>' );
+		$this->assertTrue( $tags->get_attribute( 'wanted' ), 'Found the wrong <img>' );
+	}
+
+	public function test_find_child_tag() {
+		$tags = new WP_HTML_Processor( '<div><div><div><img></div></div></div>' );
+
+		$tags->next_tag( 'div' );
+		$state              = $tags->new_state();
+		$state->match_depth = 3;
+		$this->assertTrue( $tags->balanced_next( $state, 'img' ) );
+	}
+
+	public function test_flushes_up_to_close_tag_from_deep_within() {
+		$tags = new WP_HTML_Processor(
+			<<<HTML
+			<main>
+				<section>
+					<h2>Cows</h2>
+					<div>
+						<p start>Cows are clever.</p>
+						<p>Cows eat grass.</p>
+					</div>
+					<h3>Things cows can't do</h3>
+					<ul>
+						<li><p>Pilot aeroplanes</p></li>
+						<li><p>Drive race cars</p></li>
+						<li><p>Captain ships</p></li>
+					</ul>
+					<p>This concludes our discussion of cows.</p>
+				</section>
+				<section>
+					<h2>Oxen</h2>
+					<div>
+						<p wanted>Oxen are strong.</p>
+					</div>
+				</section>
+			</main>
+HTML
+		);
+
+		$tags->next_tag( 'section' );
+		$state = $tags->new_state();
+
+		// Jump inside this tag.
+		$tags->balanced_next( $state, 'p' );
+		$this->assertTrue( $tags->get_attribute( 'start' ) );
+		// Then exit the outer section we were scanning.
+		while ( $tags->balanced_next( $state ) ) {
+			continue;
+		}
+
+		$this->assertEquals( 'SECTION', $tags->get_tag() );
+		$tags->next_tag( 'p' );
+		$this->assertTrue( $tags->get_attribute( 'wanted' ) );
+	}
+
+	public function test_can_navigate_with_unique_state_throughout_structure() {
+		$tags = new WP_HTML_Processor(
+			<<<HTML
+			<main>
+				<section>
+					<h2>Cows</h2>
+					<div>
+						<p start>Cows are clever.</p>
+						<p>Cows eat grass.</p>
+					</div>
+					<h3>Things cows can't do</h3>
+					<ul>
+						<li><p>Pilot aeroplanes</p></li>
+						<li><p>Drive race cars</p></li>
+						<li><p>Captain ships</p></li>
+					</ul>
+					<p inner>This concludes our discussion of cows.</p>
+				</section>
+				<section>
+					<h2>Oxen</h2>
+					<div>
+						<p wanted>Oxen are strong.</p>
+					</div>
+				</section>
+			</main>
+HTML
+		);
+
+		$tags->next_tag( 'section' );
+		$state = $tags->new_state();
+
+		// Jump inside this tag.
+		$tags->balanced_next( $state, 'p' );
+		$this->assertTrue( $tags->get_attribute( 'start' ) );
+
+		// Establish a new state/frame for navigating inside the outer structure.
+		$tags->balanced_next( $state, 'ul' );
+		$li_count = 0;
+		$li_state = $tags->new_state();
+		while ( $tags->balanced_next( $li_state, 'li' ) ) {
+			$li_count++;
+		}
+		$this->assertEquals( 3, $li_count );
+
+		// Ensure that we ended up where we expected.
+		$this->assertEquals( 'UL', $tags->get_tag() );
+		$this->assertTrue( $tags->is_tag_closer() );
+		$tags->next_tag();
+		$this->assertTrue( $tags->get_attribute( 'inner' ) );
+
+		// And now flush out the previous stack/frame.
+		while ( $tags->balanced_next( $state ) ) {
+			continue;
+		}
+
+		// Ensure that we're back where we want to be after exiting two separate frames.
+		$this->assertEquals( 'P', $tags->get_tag() );
+		$this->assertTrue( $tags->is_tag_closer() );
+		$tags->next_tag( 'p' );
+		$this->assertTrue( $tags->get_attribute( 'wanted' ) );
+	}
+
+	public function test_can_scan_through_tags_at_a_given_depth() {
+		$tags = new WP_HTML_Processor(
+			<<<HTML
+			<main>
+				<section>
+					<h2>Cows</h2>
+					<div>
+						<p start>Cows are clever.</p>
+						<p>Cows eat grass.</p>
+					</div>
+					<h3>Things cows can't do</h3>
+					<ul>
+						<li><p>Pilot aeroplanes</p></li>
+						<li><p>Drive race cars</p></li>
+						<li><p>Captain ships</p></li>
+					</ul>
+					<h3>Things cows can do</h3>
+					<ul>
+						<li><p>Chew cud</p></li>
+						<li><p>Moo</p></li>
+					</ul>
+					<p inner>This concludes our discussion of cows.</p>
+				</section>
+				<section>
+					<h2>Oxen</h2>
+					<div>
+						<p wanted>Oxen are strong.</p>
+					</div>
+				</section>
+			</main>
+HTML
+		);
+
+		$tags->next_tag( 'section' );
+		$state              = $tags->new_state();
+		$state->match_depth = 3;
+
+		$p3_count = 0;
+		while ( $tags->balanced_next( $state, 'p' ) ) {
+			$p3_count++;
+		}
+
+		// Did we only visit the tags inside section > * > * > p?
+		$this->assertEquals( 5, $p3_count );
+
+		$state              = $tags->new_state();
+		$state->match_depth = 2;
+
+		$p2_count = 0;
+		while ( $tags->balanced_next( $state, 'p' ) ) {
+			$p2_count++;
+		}
+
+		// Did we only visit the tags inside section > * > p?
+		$this->assertEquals( 1, $p2_count );
+	}
+}