diff --git a/pyproject.toml b/pyproject.toml
index d5514159..2e4adb1d 100644
--- a/pyproject.toml
+++ b/pyproject.toml
@@ -37,7 +37,8 @@ classifiers = [
 ]
 dependencies = [
   "awkward >=2.5.1",
-  "dask >=2023.04.0",
+  "dask >=2024.12.0;python_version>'3.9'",
+  "dask >=2023.04.0;python_version<'3.10'",
   "cachetools",
   "typing_extensions >=4.8.0",
 ]
diff --git a/src/dask_awkward/layers/__init__.py b/src/dask_awkward/layers/__init__.py
index d4ba4c5e..098bbf2e 100644
--- a/src/dask_awkward/layers/__init__.py
+++ b/src/dask_awkward/layers/__init__.py
@@ -6,6 +6,7 @@
     ImplementsIOFunction,
     ImplementsProjection,
     IOFunctionWithMocking,
+    _dask_uses_tasks,
     io_func_implements_projection,
 )
 
@@ -18,4 +19,5 @@
     "ImplementsIOFunction",
     "IOFunctionWithMocking",
     "io_func_implements_projection",
+    "_dask_uses_tasks",
 )
diff --git a/src/dask_awkward/layers/layers.py b/src/dask_awkward/layers/layers.py
index 92441443..b5c54523 100644
--- a/src/dask_awkward/layers/layers.py
+++ b/src/dask_awkward/layers/layers.py
@@ -4,6 +4,10 @@
 from collections.abc import Callable, Mapping
 from typing import TYPE_CHECKING, Any, Literal, Protocol, TypeVar, Union, cast
 
+import dask
+
+_dask_uses_tasks = hasattr(dask, "_task_spec")
+
 from dask.blockwise import Blockwise, BlockwiseDepDict, blockwise_token
 from dask.highlevelgraph import MaterializedLayer
 from dask.layers import DataFrameTreeReduction
@@ -11,6 +15,9 @@
 
 from dask_awkward.utils import LazyInputsDict
 
+if _dask_uses_tasks:
+    from dask._task_spec import Task, TaskRef
+
 if TYPE_CHECKING:
     from awkward import Array as AwkwardArray
     from awkward._nplikes.typetracer import TypeTracerReport
@@ -160,14 +167,20 @@ def __init__(
             produces_tasks=self.produces_tasks,
         )
 
-        super().__init__(
-            output=self.name,
-            output_indices="i",
-            dsk={name: (self.io_func, blockwise_token(0))},
-            indices=[(io_arg_map, "i")],
-            numblocks={},
-            annotations=None,
-        )
+        super_kwargs: dict[str, Any] = {
+            "output": self.name,
+            "output_indices": "i",
+            "indices": [(io_arg_map, "i")],
+            "numblocks": {},
+            "annotations": None,
+        }
+
+        if _dask_uses_tasks:
+            super_kwargs["task"] = Task(name, self.io_func, TaskRef(blockwise_token(0)))
+        else:
+            super_kwargs["dsk"] = {name: (self.io_func, blockwise_token(0))}
+
+        super().__init__(**super_kwargs)
 
     def __repr__(self) -> str:
         return f"AwkwardInputLayer<{self.output}>"
diff --git a/src/dask_awkward/lib/core.py b/src/dask_awkward/lib/core.py
index d7c1a4e0..786ff2bb 100644
--- a/src/dask_awkward/lib/core.py
+++ b/src/dask_awkward/lib/core.py
@@ -45,7 +45,11 @@
 from dask.utils import OperatorMethodMixin as DaskOperatorMethodMixin
 from dask.utils import funcname, is_arraylike, key_split
 
-from dask_awkward.layers import AwkwardBlockwiseLayer, AwkwardMaterializedLayer
+from dask_awkward.layers import (
+    AwkwardBlockwiseLayer,
+    AwkwardMaterializedLayer,
+    _dask_uses_tasks,
+)
 from dask_awkward.lib.optimize import all_optimizations
 from dask_awkward.utils import (
     ConcretizationTypeError,
@@ -57,6 +61,9 @@
     is_empty_slice,
 )
 
+if _dask_uses_tasks:
+    from dask._task_spec import TaskRef
+
 if TYPE_CHECKING:
     from awkward.contents.content import Content
     from awkward.forms.form import Form
@@ -1928,7 +1935,10 @@ def partitionwise_layer(
             pairs.extend([arg.name, "i"])
             numblocks[arg.name] = (1,)
         elif isinstance(arg, Delayed):
-            pairs.extend([arg.key, None])
+            if _dask_uses_tasks:
+                pairs.extend([TaskRef(arg.key), None])
+            else:
+                pairs.extend([arg.key, None])
         elif is_dask_collection(arg):
             raise DaskAwkwardNotImplemented(
                 "Use of Array with other Dask collections is currently unsupported."
diff --git a/src/dask_awkward/lib/optimize.py b/src/dask_awkward/lib/optimize.py
index 6ad2e132..4b9dd6cf 100644
--- a/src/dask_awkward/lib/optimize.py
+++ b/src/dask_awkward/lib/optimize.py
@@ -13,10 +13,17 @@
 from dask.highlevelgraph import HighLevelGraph
 from dask.local import get_sync
 
-from dask_awkward.layers import AwkwardBlockwiseLayer, AwkwardInputLayer
+from dask_awkward.layers import (
+    AwkwardBlockwiseLayer,
+    AwkwardInputLayer,
+    _dask_uses_tasks,
+)
 from dask_awkward.lib.utils import typetracer_nochecks
 from dask_awkward.utils import first
 
+if _dask_uses_tasks:
+    from dask._task_spec import GraphNode, Task, TaskRef
+
 if TYPE_CHECKING:
     from awkward._nplikes.typetracer import TypeTracerReport
     from dask.typing import Key
@@ -234,14 +241,23 @@ def _touch_all_data(*args, **kwargs):
 
 def _mock_output(layer):
     """Update a layer to run the _touch_all_data."""
-    assert len(layer.dsk) == 1
+    if _dask_uses_tasks:
+        new_layer = copy.deepcopy(layer)
+        task = new_layer.task.copy()
+        # replace the original function with _touch_all_data
+        # and keep the rest of the task the same
+        task.func = _touch_all_data
+        new_layer.task = task
+        return new_layer
+    else:
+        assert len(layer.dsk) == 1
 
-    new_layer = copy.deepcopy(layer)
-    mp = new_layer.dsk.copy()
-    for k in iter(mp.keys()):
-        mp[k] = (_touch_all_data,) + mp[k][1:]
-    new_layer.dsk = mp
-    return new_layer
+        new_layer = copy.deepcopy(layer)
+        mp = new_layer.dsk.copy()
+        for k in iter(mp.keys()):
+            mp[k] = (_touch_all_data,) + mp[k][1:]
+        new_layer.dsk = mp
+        return new_layer
 
 
 @no_type_check
@@ -340,7 +356,10 @@ def rewrite_layer_chains(dsk: HighLevelGraph, keys: Sequence[Key]) -> HighLevelG
         deps[outkey] = deps[chain[0]]
         [deps.pop(ch) for ch in chain[:-1]]
 
-        subgraph = layer0.dsk.copy()  # mypy: ignore
+        if _dask_uses_tasks:
+            all_tasks = [layer0.task]
+        else:
+            subgraph = layer0.dsk.copy()
         indices = list(layer0.indices)
         parent = chain[0]
 
@@ -349,14 +368,28 @@ def rewrite_layer_chains(dsk: HighLevelGraph, keys: Sequence[Key]) -> HighLevelG
             layer = dsk.layers[chain_member]
             for k in layer.io_deps:  # mypy: ignore
                 outlayer.io_deps[k] = layer.io_deps[k]
-            func, *args = layer.dsk[chain_member]  # mypy: ignore
-            args2 = _recursive_replace(args, layer, parent, indices)
-            subgraph[chain_member] = (func,) + tuple(args2)
+
+            if _dask_uses_tasks:
+                func = layer.task.func
+                args = [
+                    arg.key if isinstance(arg, GraphNode) else arg
+                    for arg in layer.task.args
+                ]
+                # how to do this with `.substitute(...)`?
+                args2 = _recursive_replace(args, layer, parent, indices)
+                all_tasks.append(Task(chain_member, func, *args2))
+            else:
+                func, *args = layer.dsk[chain_member]  # mypy: ignore
+                args2 = _recursive_replace(args, layer, parent, indices)
+                subgraph[chain_member] = (func,) + tuple(args2)
             parent = chain_member
         outlayer.numblocks = {
             i[0]: (numblocks,) for i in indices if i[1] is not None
         }  # mypy: ignore
-        outlayer.dsk = subgraph  # mypy: ignore
+        if _dask_uses_tasks:
+            outlayer.task = Task.fuse(*all_tasks)
+        else:
+            outlayer.dsk = subgraph  # mypy: ignore
         if hasattr(outlayer, "_dims"):
             del outlayer._dims
         outlayer.indices = tuple(  # mypy: ignore
@@ -379,11 +412,18 @@ def _recursive_replace(args, layer, parent, indices):
                 args2.append(layer.indices[ind][0])
             elif layer.indices[ind][0] == parent:
                 # arg refers to output of previous layer
-                args2.append(parent)
+                if _dask_uses_tasks:
+                    args2.append(TaskRef(parent))
+                else:
+                    args2.append(parent)
             else:
                 # arg refers to things defined in io_deps
                 indices.append(layer.indices[ind])
-                args2.append(f"__dask_blockwise__{len(indices) - 1}")
+                arg2 = f"__dask_blockwise__{len(indices) - 1}"
+                if _dask_uses_tasks:
+                    args2.append(TaskRef(arg2))
+                else:
+                    args2.append(arg2)
         elif isinstance(arg, list):
             args2.append(_recursive_replace(arg, layer, parent, indices))
         elif isinstance(arg, tuple):
diff --git a/tests/test_io_json.py b/tests/test_io_json.py
index 688fb550..f7192aae 100644
--- a/tests/test_io_json.py
+++ b/tests/test_io_json.py
@@ -10,6 +10,7 @@
 import pytest
 
 import dask_awkward as dak
+from dask_awkward.layers import _dask_uses_tasks
 from dask_awkward.lib.core import Array
 from dask_awkward.lib.optimize import optimize as dak_optimize
 from dask_awkward.lib.testutils import assert_eq
@@ -94,8 +95,11 @@ def input_layer_array_partition0(collection: Array) -> ak.Array:
         optimized_hlg = dak_optimize(collection.dask, collection.keys)  # type: ignore
         layers = list(optimized_hlg.layers)  # type: ignore
         layer_name = [name for name in layers if name.startswith("from-json")][0]
-        sgc, arg = optimized_hlg[(layer_name, 0)]
-        array = sgc.dsk[layer_name][0](arg)
+        if _dask_uses_tasks:
+            array = optimized_hlg[(layer_name, 0)]()
+        else:
+            sgc, arg = optimized_hlg[(layer_name, 0)]
+            array = sgc.dsk[layer_name][0](arg)
     return array