diff --git a/ome_zarr/io.py b/ome_zarr/io.py
index c3ad9de5..67ab246e 100644
--- a/ome_zarr/io.py
+++ b/ome_zarr/io.py
@@ -6,15 +6,23 @@
 import json
 import logging
 from pathlib import Path
-from typing import List, Optional, Union
+from typing import Callable, List, Optional, Union
 from urllib.parse import urljoin
 
 import dask.array as da
+import zarr as zarrlib
+from anndata import AnnData
+from anndata._io.specs import IOSpec
+from anndata.compat import H5Array, H5Group, ZarrArray, ZarrGroup
+
+# ** requires anndata==0.9.0.rc1
+from anndata.experimental import read_dispatched, read_elem
 from zarr.storage import FSStore
 
 from .format import CurrentFormat, Format, detect_format
 from .types import JSONDict
 
+StorageType = Union[H5Array, H5Group, ZarrArray, ZarrGroup]
 LOGGER = logging.getLogger("ome_zarr.io")
 
 
@@ -194,3 +202,30 @@ def parse_url(
         LOGGER.exception("exception on parsing (stacktrace at DEBUG)")
         LOGGER.debug("stacktrace:", exc_info=True)
         return None
+
+
+# From https://anndata.readthedocs.io/en/latest/tutorials/
+# notebooks/%7Bread,write%7D_dispatched.html
+
+
+def read_remote_anndata(store: FSStore, name: str) -> AnnData:
+    table_group = zarrlib.group(store=store, path=name)
+
+    def callback(
+        func: Callable, elem_name: str, elem: StorageType, iospec: IOSpec
+    ) -> AnnData:
+        if iospec.encoding_type in (
+            "dataframe",
+            "csr_matrix",
+            "csc_matrix",
+            "awkward-array",
+        ):
+            # Preventing recursing inside of these types
+            return read_elem(elem)
+        elif iospec.encoding_type == "array":
+            return da.from_zarr(elem)
+        else:
+            return func(elem)
+
+    adata = read_dispatched(table_group, callback=callback)
+    return adata
diff --git a/ome_zarr/reader.py b/ome_zarr/reader.py
index 3592a358..b13dd896 100644
--- a/ome_zarr/reader.py
+++ b/ome_zarr/reader.py
@@ -2,22 +2,17 @@
 
 import logging
 import math
-import os
 from abc import ABC
 from typing import Any, Dict, Iterator, List, Optional, Type, Union, cast, overload
 
 import dask.array as da
 import numpy as np
-
-# experimental failed to import
-# from anndata.experimental import read_dispatched, write_dispatched, read_elem
 from anndata import AnnData
-from anndata._io import read_zarr
 from dask import delayed
 
 from .axes import Axes
 from .format import format_from_version
-from .io import ZarrLocation
+from .io import ZarrLocation, read_remote_anndata
 from .types import JSONDict
 
 LOGGER = logging.getLogger("ome_zarr.reader")
@@ -187,32 +182,6 @@ def lookup(self, key: str, default: Any) -> Any:
         return self.zarr.root_attrs.get(key, default)
 
 
-# From https://anndata.readthedocs.io/en/latest/tutorials/
-# notebooks/%7Bread,write%7D_dispatched.html
-
-# def read_dask(store, path):
-#     f = zarr.open(store, path=path, mode="r")
-
-#     def callback(func, elem_name: str, elem, iospec):
-#         print("callback", iospec.encoding_type)
-#         if iospec.encoding_type in (
-#             "dataframe",
-#             "csr_matrix",
-#             "csc_matrix",
-#             "awkward-array",
-#         ):
-#             # Preventing recursing inside of these types
-#             return read_elem(elem)
-#         elif iospec.encoding_type == "array":
-#             return da.from_zarr(elem)
-#         else:
-#             return func(elem)
-
-#     adata = read_dispatched(f, callback=callback)
-
-#     return adata
-
-
 class Tables(Spec):
     """Class to represent a "tables" group which only
     contains the name of subgroups which should be loaded as labeled images."""
@@ -226,15 +195,13 @@ def __init__(self, node: Node) -> None:
         super().__init__(node)
         table_names = self.lookup("tables", [])
         node.tables = {}
-        store = self.zarr.store
         for name in table_names:
             child_zarr = self.zarr.create(name)
             if child_zarr.exists():
                 node.add(child_zarr)
-
-                # node.tables[name] = read_dask(store, name)
-                full_path = os.path.join(store.path, name)
-                node.tables[name] = read_zarr(full_path)
+                LOGGER.info("Reading anndata table: %s", name)
+                anndata_obj = read_remote_anndata(node.zarr.store, name)
+                node.tables[name] = anndata_obj
 
 
 class Labels(Spec):
diff --git a/setup.py b/setup.py
index 940870ad..2deca46e 100644
--- a/setup.py
+++ b/setup.py
@@ -13,7 +13,7 @@ def read(fname):
 
 
 install_requires: List[List[str]] = []
-install_requires += (["anndata"],)
+install_requires += (["anndata=0.9.0.rc1"],)
 install_requires += (["dataclasses;python_version<'3.7'"],)
 install_requires += (["tifffile<2020.09.22;python_version<'3.7'"],)
 install_requires += (["numpy"],)