diff --git a/src/mkdocstrings_handlers/python/_internal/handler.py b/src/mkdocstrings_handlers/python/_internal/handler.py
index 158d7dd..3246919 100644
--- a/src/mkdocstrings_handlers/python/_internal/handler.py
+++ b/src/mkdocstrings_handlers/python/_internal/handler.py
@@ -228,9 +228,9 @@ def collect(self, identifier: str, options: PythonOptions) -> CollectorItem:
         parser = parser_name and Parser(parser_name)
         parser_options = options.docstring_options and asdict(options.docstring_options)
 
-        if unknown_module:
+        def make_loader():
             extensions = self.normalize_extension_paths(options.extensions)
-            loader = GriffeLoader(
+            return GriffeLoader(
                 extensions=load_extensions(*extensions),
                 search_paths=self._paths,
                 docstring_parser=parser,
@@ -240,6 +240,10 @@ def collect(self, identifier: str, options: PythonOptions) -> CollectorItem:
                 allow_inspection=options.allow_inspection,
                 force_inspection=options.force_inspection,
             )
+
+        loader = None
+        if unknown_module:
+            loader = make_loader()
             try:
                 for pre_loaded_module in options.preload_modules:
                     if pre_loaded_module not in self._modules_collection:
@@ -263,6 +267,11 @@ def collect(self, identifier: str, options: PythonOptions) -> CollectorItem:
                 _logger.debug(f"{len(unresolved)} aliases were still unresolved after {iterations} iterations")
                 _logger.debug(f"Unresolved aliases: {', '.join(sorted(unresolved))}")
 
+        if identifier not in self._modules_collection:
+            if loader is None:
+                loader = make_loader()
+            self.try_to_import(identifier, loader)
+
         try:
             doc_object = self._modules_collection[identifier]
         except KeyError as error:
@@ -278,6 +287,58 @@ def collect(self, identifier: str, options: PythonOptions) -> CollectorItem:
 
         return doc_object
 
+
+    def try_to_import(self, identifier: str, loader):
+        import importlib.util
+        from pathlib import Path
+
+        parts = identifier.split(".")
+        origins = [None] * len(parts)
+        child_origin = None
+        for i in range(len(parts) - 1, -1, -1):
+            sub_identifier = ".".join(parts[:i + 1])
+            spec = importlib.util.find_spec(sub_identifier)
+            if spec is None:
+                continue
+            origin = spec.origin
+            if origin is None and child_origin is not None:
+                # Pick the first loc that is a parent of child_origin
+                origin = next(
+                    (
+                        loc
+                        for loc in spec.submodule_search_locations
+                        if child_origin.startswith(loc)
+                    ),
+                    None,
+                )
+            origins[i] = origin
+            if origin is not None:
+                child_origin = origin
+
+        parts = ()
+        doc_object = None
+        for i, part in enumerate(identifier.split(".")):
+            parts = parts + (part,)
+            sub_identifier = ".".join(parts)
+            origin = origins[i]
+            if origin is None:
+                if doc_object is not None and part in doc_object.members:
+                    # If we already have a doc_object, we can just use it
+                    doc_object = doc_object.get_member(part)
+                    continue
+                raise CollectionError(f"Module {identifier} could not be found")
+
+            doc_object = loader._load_module(
+                part,
+                Path(origin),
+                parent=doc_object,
+            )
+            try:
+                self._modules_collection.set_member(sub_identifier, doc_object)
+            except:
+                pass
+        return doc_object
+
     def render(self, data: CollectorItem, options: PythonOptions) -> str:
         """Render the collected data.
 

<!DOCTYPE html PUBLIC '-//W3C//DTD XHTML 1.0 Transitional//EN' 'http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd'>
<html xmlns='http://www.w3.org/1999/xhtml'>
<head>
<title>pFad - Phonifier reborn</title>
<meta http-equiv='Content-Type' content='text/html; charset=utf-8' />
</head>
<body>
<h1>Pfad - The Proxy pFad of &#169; 2024 Garber Painting. All rights reserved.</h1>


<!-- Disclaimer -->
<p>Note: This service is not intended for secure transactions such as banking, social media, email, or purchasing. Use at your own risk. We assume no liability whatsoever for broken pages.</p>
<br>
<p>Alternative Proxies:</p><p><a href="http://rainy.clevelandohioweatherforecast.com/php-proxy/index.php?q=https://github.com/mkdocstrings/python/compare/main...percevalw:mkdocstrings-python:main.diff" target="_blank">Alternative Proxy</a></p><p><a href="http://rainy.clevelandohioweatherforecast.com/pFad/index.php?u=https://github.com/mkdocstrings/python/compare/main...percevalw:mkdocstrings-python:main.diff" target="_blank">pFad Proxy</a></p><p><a href="http://rainy.clevelandohioweatherforecast.com/pFad/v3index.php?u=https://github.com/mkdocstrings/python/compare/main...percevalw:mkdocstrings-python:main.diff" target="_blank">pFad v3 Proxy</a></p><p><a href="http://rainy.clevelandohioweatherforecast.com/pFad/v4index.php?u=https://github.com/mkdocstrings/python/compare/main...percevalw:mkdocstrings-python:main.diff" target="_blank">pFad v4 Proxy</a></p></body>
</html>