diff --git a/lib/iris/fileformats/__init__.py b/lib/iris/fileformats/__init__.py
index 0af0888b3f..8de17eb5b8 100644
--- a/lib/iris/fileformats/__init__.py
+++ b/lib/iris/fileformats/__init__.py
@@ -155,6 +155,20 @@ def _load_grib(*args, **kwargs):
 )
 
 
+#
+# Zarr files.
+#
+FORMAT_AGENT.add_spec(
+    FormatSpecification(
+        "zarr",
+        FileExtension(requires_fh=False),
+        ".zarr#mode=nczarr,file",
+        netcdf.load_cubes,
+        priority=3,
+    )
+)
+
+
 #
 # UM Fieldsfiles.
 #
diff --git a/lib/iris/fileformats/netcdf/saver.py b/lib/iris/fileformats/netcdf/saver.py
index 98c565e990..fc420a2a99 100644
--- a/lib/iris/fileformats/netcdf/saver.py
+++ b/lib/iris/fileformats/netcdf/saver.py
@@ -418,7 +418,11 @@ def __init__(self, filename, netcdf_format, compute=True):
         else:
             # Given a filepath string/path : create a dataset from that
             try:
-                self.filepath = Path(filename).absolute()
+                filepath = Path(filename)
+                if ".zarr" not in filepath.suffix:
+                    self.filepath = filepath.absolute()
+                else:
+                    self.filepath = filepath
                 self._dataset = _thread_safe_nc.DatasetWrapper(
                     self.filepath, mode="w", format=netcdf_format
                 )
diff --git a/lib/iris/io/__init__.py b/lib/iris/io/__init__.py
index 1a895ff1a3..3a6f38c44d 100644
--- a/lib/iris/io/__init__.py
+++ b/lib/iris/io/__init__.py
@@ -129,7 +129,10 @@ def decode_uri(uri, default="file"):
         # put - last in the brackets so it refers to the character, not a range
         # reference on valid schemes: https://tools.ietf.org/html/std66#section-3.1
         match = re.match(r"^([a-zA-Z][a-zA-Z0-9+.-]+):(.+)", uri)
-        if match:
+        if ".zarr" in uri:
+            scheme = "zarr"
+            part = uri
+        elif match:
             scheme = match.group(1)
             part = match.group(2)
         else:
diff --git a/lib/iris/loading.py b/lib/iris/loading.py
index 68042847c1..69bfab30e3 100644
--- a/lib/iris/loading.py
+++ b/lib/iris/loading.py
@@ -97,6 +97,10 @@ def _generate_cubes(uris, callback, constraints):
             urls = [":".join(x) for x in groups]
             for cube in iris.io.load_http(urls, callback):
                 yield cube
+        elif scheme == "zarr":
+            urls = [x[1] for x in groups]
+            for cube in iris.io.load_http(urls, callback):
+                yield cube
         elif scheme == "data":
             data_objects = [x[1] for x in groups]
             for cube in iris.io.load_data_objects(data_objects, callback):
diff --git a/lib/iris/tests/results/file_load/known_loaders.txt b/lib/iris/tests/results/file_load/known_loaders.txt
index 98ac3e4a07..8aa7afc4e1 100644
--- a/lib/iris/tests/results/file_load/known_loaders.txt
+++ b/lib/iris/tests/results/file_load/known_loaders.txt
@@ -14,4 +14,5 @@
  * UM Fieldsfile (FF) converted with ieee to 32 bit (priority 3)
  * UM Fieldsfile (FF) pre v3.1 (priority 3)
  * UM Post Processing file (PP) little-endian (priority 3)
+ * zarr (priority 3)
  * GRIB (priority 1)
\ No newline at end of file