diff --git a/CHANGELOG.md b/CHANGELOG.md
index 8fc818f7..860a8851 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -1,6 +1,8 @@
 # 0.0.5 2024-04-XX
 
 - Fix bug in schema handling (compressor settings ignored)
+- Move making ICF field partition directories into per-partition processing.
+  Remove progress on the init mkdirs step.
 
 # 0.0.4 2024-04-08
 
diff --git a/bio2zarr/vcf.py b/bio2zarr/vcf.py
index 1889279c..10bbadd6 100644
--- a/bio2zarr/vcf.py
+++ b/bio2zarr/vcf.py
@@ -791,6 +791,8 @@ def __init__(
         for vcf_field in icf_metadata.fields:
             field_path = get_vcf_field_path(out_path, vcf_field)
             field_partition_path = field_path / f"p{partition_index}"
+            # Should be robust to running explode_partition twice.
+            field_partition_path.mkdir(exist_ok=True)
             transformer = VcfValueTransformer.factory(vcf_field, num_samples)
             self.field_writers[vcf_field.full_name] = IcfFieldWriter(
                 vcf_field,
@@ -890,14 +892,6 @@ def num_columns(self):
         return len(self.columns)
 
 
-def mkdir_with_progress(path):
-    logger.debug(f"mkdir f{path}")
-    # NOTE we may have race-conditions here, I'm not sure. Hopefully allowing
-    # parents=True will take care of it.
-    path.mkdir(parents=True)
-    core.update_progress(1)
-
-
 class IntermediateColumnarFormatWriter:
     def __init__(self, path):
         self.path = pathlib.Path(path)
@@ -940,7 +934,7 @@ def init(
         # dependencies as well.
         self.metadata.provenance = {"source": f"bio2zarr-{provenance.__version__}"}
 
-        self.mkdirs(worker_processes, show_progress=show_progress)
+        self.mkdirs()
 
         # Note: this is needed for the current version of the vcfzarr spec, but it's
         # probably going to be dropped.
@@ -955,30 +949,14 @@ def init(
             json.dump(self.metadata.asdict(), f, indent=4)
         return self.num_partitions
 
-    def mkdirs(self, worker_processes=1, show_progress=False):
-        num_dirs = len(self.metadata.fields) * self.num_partitions
-        logger.info(f"Creating {num_dirs} directories")
+    def mkdirs(self):
+        num_dirs = len(self.metadata.fields)
+        logger.info(f"Creating {num_dirs} field directories")
         self.path.mkdir()
         self.wip_path.mkdir()
-        # Due to high latency batch system filesystems, we create all the directories in
-        # parallel
-        progress_config = core.ProgressConfig(
-            total=num_dirs,
-            units="dirs",
-            title="Mkdirs",
-            show=show_progress,
-        )
-        with core.ParallelWorkManager(
-            worker_processes=worker_processes, progress_config=progress_config
-        ) as manager:
-            for field in self.metadata.fields:
-                col_path = get_vcf_field_path(self.path, field)
-                # Don't bother trying to count the intermediate directories towards
-                # progress
-                manager.submit(col_path.mkdir, parents=True)
-                for j in range(self.num_partitions):
-                    part_path = col_path / f"p{j}"
-                    manager.submit(mkdir_with_progress, part_path)
+        for field in self.metadata.fields:
+            col_path = get_vcf_field_path(self.path, field)
+            col_path.mkdir(parents=True)
 
     def load_partition_summaries(self):
         summaries = []