diff --git a/README.md b/README.md
index e19127d..50f5be0 100644
--- a/README.md
+++ b/README.md
@@ -236,10 +236,11 @@ For ASGI servers:
 - [Uvicorn](https://www.uvicorn.org/) - Lightning-fast ASGI server
 - [Daphne](https://github.com/django/daphne) - Django Channels' ASGI server with HTTP/2 support
 - [Hypercorn](https://gitlab.com/pgjones/hypercorn) - ASGI server with HTTP/2 and HTTP/3 support
+- [Gunicorn](https://gunicorn.org/) - Python WSGI and ASGI HTTP Server
 
 For WSGI servers:
 
-- [Gunicorn](https://gunicorn.org/) - Python WSGI HTTP Server
+- [Gunicorn](https://gunicorn.org/) - Python WSGI and ASGI HTTP Server
 - [uWSGI](https://uwsgi-docs.readthedocs.io/) - Full-featured application server
 - Any WSGI-compliant server
 
diff --git a/conformance/test/server.py b/conformance/test/server.py
index ef39ecc..27f28ae 100644
--- a/conformance/test/server.py
+++ b/conformance/test/server.py
@@ -543,18 +543,17 @@ async def serve_granian(
 
 async def serve_gunicorn(
     request: ServerCompatRequest,
+    mode: Mode,
     certfile: str | None,
     keyfile: str | None,
     cafile: str | None,
     port_future: asyncio.Future[int],
 ):
-    args = [
-        "--bind=127.0.0.1:0",
-        "--workers=4",
-        "--worker-class=gthread",
-        "--threads=16",
-        "--keep-alive=0",
-    ]
+    args = ["--bind=127.0.0.1:0", "--workers=4"]
+    if mode == "sync":
+        args.extend(["--worker-class=gthread", "--threads=16", "--keep-alive=0"])
+    else:
+        args.append("--worker-class=asgi")
     if certfile:
         args.append(f"--certfile={certfile}")
     if keyfile:
@@ -563,7 +562,7 @@ async def serve_gunicorn(
         args.append(f"--ca-certs={cafile}")
         args.append(f"--cert-reqs={ssl.CERT_REQUIRED}")
 
-    args.append("server:wsgi_app")
+    args.append("server:wsgi_app" if mode == "sync" else "server:asgi_app")
 
     proc = await asyncio.create_subprocess_exec(
         "gunicorn",
@@ -791,11 +790,10 @@ async def main() -> None:
                     )
                 )
             case "gunicorn":
-                if args.mode == "async":
-                    msg = "gunicorn does not support async mode"
-                    raise ValueError(msg)
                 serve_task = asyncio.create_task(
-                    serve_gunicorn(request, certfile, keyfile, cafile, port_future)
+                    serve_gunicorn(
+                        request, args.mode, certfile, keyfile, cafile, port_future
+                    )
                 )
             case "hypercorn":
                 serve_task = asyncio.create_task(
diff --git a/conformance/test/test_server.py b/conformance/test/test_server.py
index a38ac35..0d66c43 100644
--- a/conformance/test/test_server.py
+++ b/conformance/test/test_server.py
@@ -42,6 +42,8 @@ def macos_raise_ulimit():
     "Server Message Size/HTTPVersion:1/**/first-request-exceeds-server-limit",
 ]
 
+_skip_http2_http3 = ["--skip", "**/HTTPVersion:2/**", "--skip", "**/HTTPVersion:3/**"]
+
 
 @pytest.mark.parametrize("server", ["gunicorn", "pyvoy"])
 def test_server_sync(server: str, cov: Coverage) -> None:
@@ -52,7 +54,7 @@ def test_server_sync(server: str, cov: Coverage) -> None:
     match server:
         case "gunicorn":
             # gunicorn doesn't support HTTP/2 or 3
-            opts = ["--skip", "**/HTTPVersion:2/**", "--skip", "**/HTTPVersion:3/**"]
+            opts = _skip_http2_http3
 
     result = subprocess.run(
         [
@@ -77,7 +79,7 @@ def test_server_sync(server: str, cov: Coverage) -> None:
         pytest.fail(f"\n{result.stdout}\n{result.stderr}")
 
 
-@pytest.mark.parametrize("server", ["daphne", "pyvoy", "uvicorn"])
+@pytest.mark.parametrize("server", ["daphne", "gunicorn", "pyvoy", "uvicorn"])
 def test_server_async(server: str, cov: Coverage) -> None:
     args = maybe_patch_args_with_debug(
         [sys.executable, _server_py_path, "--mode", "async", "--server", server]
@@ -104,9 +106,24 @@ def test_server_async(server: str, cov: Coverage) -> None:
                 "--skip",
                 "gRPC Unexpected Requests/**",
             ]
+        case "gunicorn":
+            opts = [
+                # gunicorn's native HTTP/2 ASGI worker sends GOAWAY after very few streams
+                # under load; no upstream issue filed yet (see PR #3568 for related h2 fix)
+                *_skip_http2_http3,
+                # gunicorn's gunicorn_h1c C parser returns 400 "Invalid request line" for
+                # gRPC-Web requests broadly; no upstream issue filed yet
+                # (see issue #3563 for a related gunicorn_h1c strictness regression)
+                "--skip",
+                "**/Protocol:PROTOCOL_GRPC_WEB/**",
+                "--skip",
+                "gRPC-Web Proto Sub-Format Requests/**",
+                "--skip",
+                "gRPC-Web Unexpected Requests/**",
+            ]
         case "uvicorn":
             # uvicorn doesn't support HTTP/2 or 3
-            opts = ["--skip", "**/HTTPVersion:2/**", "--skip", "**/HTTPVersion:3/**"]
+            opts = _skip_http2_http3
     result = subprocess.run(
         [
             "go",