diff --git a/python/cugraph/cugraph/experimental/__init__.py b/python/cugraph/cugraph/experimental/__init__.py
index 1c24389bafc..3ecb9723887 100644
--- a/python/cugraph/cugraph/experimental/__init__.py
+++ b/python/cugraph/cugraph/experimental/__init__.py
@@ -27,4 +27,6 @@
 
 from cugraph.experimental.community.triangle_count import \
     EXPERIMENTAL__triangle_count
-triangle_count = experimental_warning_wrapper(EXPERIMENTAL__triangle_count)
\ No newline at end of file
+triangle_count = experimental_warning_wrapper(EXPERIMENTAL__triangle_count)
+
+from cugraph.experimental.datasets.dataset import Dataset
diff --git a/python/cugraph/cugraph/experimental/datasets/__init__.py b/python/cugraph/cugraph/experimental/datasets/__init__.py
new file mode 100644
index 00000000000..675961975b1
--- /dev/null
+++ b/python/cugraph/cugraph/experimental/datasets/__init__.py
@@ -0,0 +1,36 @@
+# Copyright (c) 2022, NVIDIA CORPORATION.
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+#from cugraph.experimental.datasets import (
+#    karate,
+#)
+from cugraph.experimental.datasets.dataset import (
+    Dataset,
+#    karate,
+#    dolphins,
+#    SMALL_DATASETS
+)
+from cugraph.experimental.datasets import metadata
+
+# SMALL DATASETS
+karate = Dataset("metadata/karate.yaml")
+dolphins = Dataset("metadata/dolphins.yaml")
+polbooks = Dataset("metadata/polbooks.yaml")
+netscience = Dataset("metadata/netscience.yaml")
+
+# MEDIUM DATASETS
+
+# LARGE DATASETS
+
+# GROUPS OF DATASETS
+SMALL_DATASETS = [karate, dolphins]
\ No newline at end of file
diff --git a/python/cugraph/cugraph/experimental/datasets/dataset.py b/python/cugraph/cugraph/experimental/datasets/dataset.py
new file mode 100644
index 00000000000..dbaf9e63986
--- /dev/null
+++ b/python/cugraph/cugraph/experimental/datasets/dataset.py
@@ -0,0 +1,115 @@
+# Copyright (c) 2022, NVIDIA CORPORATION.
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+import cugraph
+import cudf
+import yaml
+import requests
+import re
+import os
+import csv
+import pdb
+
+class Dataset:
+    def __init__(self, meta_data_file_name):
+        self.dir_path = "python/cugraph/cugraph/experimental/datasets/"
+        self.download_dir = "datasets/"
+        self.__read_config()
+        self.__meta_data_file_name = meta_data_file_name
+        self.__read_meta_data_file(self.__meta_data_file_name)
+        self.__edgelist = None
+        self.__graph = None
+
+
+    def __read_meta_data_file(self, meta_data_file):
+        with open(self.dir_path + meta_data_file, 'r') as file:
+            self.metadata = yaml.safe_load(file)
+            file.close()
+
+    
+    def __read_config(self):
+        config_path = "python/cugraph/cugraph/experimental/datasets/datasets_config.yaml"
+        with open(config_path, 'r') as file:
+            cfg = yaml.safe_load(file)
+            self.download_dir = cfg['download_dir'] # should this be accessible by user?
+            file.close()
+
+
+    def __download_csv(self, url, default_path):
+        filename = url.split('/')[-1]
+        df = cudf.read_csv(url)
+        df.to_csv(default_path+filename, index=False)
+        self.metadata['path'] = default_path + filename
+
+        with open(self.dir_path + self.__meta_data_file_name, 'w') as file:
+            yaml.dump(self.metadata, file, sort_keys=False)
+            file.close()
+
+    # figure out throwing errors if fetch=False and file doesn't exist...
+    def get_edgelist(self, fetch=False):
+        """
+        Return an Edgelist
+
+        Parameters
+        ----------
+        fetch : Boolean (default=False)
+        
+        Automatically fetch for the dataset from the 'url' location within the YAML file. 
+        """
+        if self.__edgelist is None:
+            if not os.path.isfile(self.metadata['path']):
+                if fetch:
+                    self.__download_csv(self.metadata['url'], self.download_dir)
+                else:
+                    raise RuntimeError("The datafile does not exist. Try get_edgelist(fetch=True) to download the datafile")
+    
+            self.__edgelist = cudf.read_csv(self.metadata['path'], delimiter=self.metadata['delimiter'], names=self.metadata['col_names'], dtype=self.metadata['col_types'])
+
+        return self.__edgelist
+
+
+    def get_graph(self, fetch=False):
+        """
+        Return a Graph object.
+
+        Parameters
+        ----------
+        fetch : Boolean (default=False)
+        
+        Automatically fetch for the dataset from the 'url' location within the YAML file. 
+        """
+        if self.__edgelist is None:
+            self.get_edgelist(fetch)
+
+        self.__graph = cugraph.Graph(directed=self.metadata['is_directed'])
+        self.__graph.from_cudf_edgelist(self.__edgelist, source='src', destination='dst')
+
+        return self.__graph
+
+
+    def load_all(self):
+        """
+        Looks in `metadata` directory and fetches all datafiles from the web.
+        """
+        meta_path = self.dir_path + 'metadata/'
+        for file in os.listdir(meta_path):
+            print(file)
+            meta = None
+            #pdb.set_trace()
+            if file.split('.')[-1] == 'yaml':
+                with open(meta_path + file, 'r') as metafile:
+                    meta = yaml.safe_load(metafile)
+                    metafile.close()
+
+                print("downloading from " + meta['url'])
+                self.__download_csv(meta['url'], self.download_dir)
diff --git a/python/cugraph/cugraph/experimental/datasets/datasets_config.yaml b/python/cugraph/cugraph/experimental/datasets/datasets_config.yaml
new file mode 100644
index 00000000000..bbc3de17129
--- /dev/null
+++ b/python/cugraph/cugraph/experimental/datasets/datasets_config.yaml
@@ -0,0 +1,3 @@
+---
+fetch: "False"
+download_dir: "datasets2/"
diff --git a/python/cugraph/cugraph/experimental/datasets/metadata/__init__.py b/python/cugraph/cugraph/experimental/datasets/metadata/__init__.py
new file mode 100644
index 00000000000..e69de29bb2d
diff --git a/python/cugraph/cugraph/experimental/datasets/metadata/dolphins.yaml b/python/cugraph/cugraph/experimental/datasets/metadata/dolphins.yaml
new file mode 100644
index 00000000000..81ac15a9506
--- /dev/null
+++ b/python/cugraph/cugraph/experimental/datasets/metadata/dolphins.yaml
@@ -0,0 +1,22 @@
+name: dolphins
+author: D. Lusseau
+url: https://raw.githubusercontent.com/rapidsai/cugraph/branch-22.08/datasets/dolphins.csv
+refs: D. Lusseau, K. Schneider, O. J. Boisseau, P. Haase, E. Slooten, and S. M. Dawson,
+  The bottlenose dolphin community of Doubtful Sound features a large proportion of
+  long-lasting associations, Behavioral Ecology and Sociobiology 54, 396-405 (2003).
+col_names:
+- src
+- dst
+- wgt
+col_types:
+- int32
+- int32
+- float32
+delimiter: ' '
+has_loop: true
+is_directed: true
+is_multigraph: false
+is_symmetric: false
+number_of_edges: 159
+number_of_nodes: 62
+path: python/cugraph/dolphins.csv
diff --git a/python/cugraph/cugraph/experimental/datasets/metadata/karate.yaml b/python/cugraph/cugraph/experimental/datasets/metadata/karate.yaml
new file mode 100644
index 00000000000..2740e33c845
--- /dev/null
+++ b/python/cugraph/cugraph/experimental/datasets/metadata/karate.yaml
@@ -0,0 +1,19 @@
+name: karate
+author: Zachary W.
+url: https://raw.githubusercontent.com/rapidsai/cugraph/branch-22.08/datasets/karate-data.csv
+refs: W. W. Zachary, An information flow model for conflict and fission in small groups,
+  Journal of Anthropological Research 33, 452-473 (1977).
+delimiter: "\t"
+col_names:
+- src
+- dst
+col_types:
+- int32
+- int32
+has_loop: true
+is_directed: true
+is_multigraph: false
+is_symmetric: true
+number_of_edges: 156
+number_of_nodes: 34
+path: datasets2/dolphins.csv
diff --git a/python/cugraph/cugraph/experimental/datasets/metadata/netscience.yaml b/python/cugraph/cugraph/experimental/datasets/metadata/netscience.yaml
new file mode 100644
index 00000000000..5de8acd232b
--- /dev/null
+++ b/python/cugraph/cugraph/experimental/datasets/metadata/netscience.yaml
@@ -0,0 +1,20 @@
+name: netscience
+author: Newman, Mark EJ
+url: https://raw.githubusercontent.com/rapidsai/cugraph/branch-22.08/datasets/netscience.csv
+refs: Finding community structure in networks using the eigenvectors of matrices.
+delimiter: " "
+col_names:
+  - src
+  - dst
+  - wgt
+col_types:
+  - int32
+  - int32
+  - float32
+has_loop: false
+is_directed: false
+is_multigraph: false
+is_symmetric: true
+number_of_edges: 441
+number_of_nodes: 105
+path: datasets/netscience.csv
diff --git a/python/cugraph/cugraph/experimental/datasets/metadata/polbooks.yaml b/python/cugraph/cugraph/experimental/datasets/metadata/polbooks.yaml
new file mode 100644
index 00000000000..c0e310f3626
--- /dev/null
+++ b/python/cugraph/cugraph/experimental/datasets/metadata/polbooks.yaml
@@ -0,0 +1,18 @@
+name: polbooks
+author: V. Krebs
+url: https://raw.githubusercontent.com/rapidsai/cugraph/branch-22.08/datasets/polbooks.csv
+refs: null
+delimiter: null
+col_names:
+- src
+- dst
+col_types:
+- int32
+- int32
+has_loop: null
+is_directed: false
+is_multigraph: null
+is_symmetric: true
+number_of_edges: 441
+number_of_nodes: 105
+path: datasets2/polbooks.csv
diff --git a/python/cugraph/cugraph/tests/test_dataset.py b/python/cugraph/cugraph/tests/test_dataset.py
new file mode 100644
index 00000000000..c55459df2f3
--- /dev/null
+++ b/python/cugraph/cugraph/tests/test_dataset.py
@@ -0,0 +1,42 @@
+# Copyright (c) 2022, NVIDIA CORPORATION.
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+import gc
+
+import pytest
+import cugraph
+from cugraph.experimental.datasets import dataset, SMALL_DATASETS
+
+# =============================================================================
+# Pytest Setup / Teardown - called for each test function
+# =============================================================================
+def setup_function():
+    gc.collect()
+
+# A simple example Dataset class working, MetaData
+# run thru an entire alg with imports
+# handle cases, like fetch, also maybe config options
+#config_file_path = "cugraph/cugraph/experimental/datasets/datasets_config.yaml"
+#with open(config_file_path, 'r') as file:
+#    config_settings = yaml.safe_load(file)
+
+@pytest.mark.parametrize("dataset", SMALL_DATASETS)
+def test_getters(dataset):
+    # Getting the graph does not need to depend on get_edgelist
+    M = dataset.get_edgelist(fetch=True)
+    #breakpoint()
+    G = dataset.get_graph(fetch=True)
+
+    # Storing the datasets in experimental/datasets/
+
+# Test that no fetches are redundant; i.e if dataset has been fetched or already exists, don't fetch again