diff --git a/setup.py b/setup.py
index 496604bcc..d539a2c20 100755
--- a/setup.py
+++ b/setup.py
@@ -59,6 +59,7 @@
 scripts = ['src/python/pcmdi/scripts/mean_climate_driver.py',
            'src/python/pcmdi/scripts/pcmdi_compute_climatologies.py',
            'src/python/misc/scripts/parallelize_driver.py',
+           'src/python/misc/scripts/get_pmp_data.py',
            'src/python/monsoon_wang/scripts/mpindex_compute.py',
            'src/python/monsoon_sperber/scripts/driver_monsoon_sperber.py',
            ]
diff --git a/src/python/misc/scripts/get_pmp_data.py b/src/python/misc/scripts/get_pmp_data.py
new file mode 100644
index 000000000..3c8747caf
--- /dev/null
+++ b/src/python/misc/scripts/get_pmp_data.py
@@ -0,0 +1,50 @@
+#!/usr/bin/env python
+from __future__ import print_function
+from pcmdi_metrics.driver.pmp_parser import PMPParser
+import tempfile
+import requests
+import os
+import cdat_info
+
+
+def download_file(download_url_root, name, local_filename):
+    r = requests.get("%s/%s" % (download_url_root, name), stream=True)
+    with open(local_filename, 'wb') as f:
+        for chunk in r.iter_content(chunk_size=1024):
+            if chunk:  # filter local_filename keep-alive new chunks
+                f.write(chunk)
+
+
+parser = PMPParser(description='Get sample data')
+parser.add_argument("--dataset", help="Download observation or sample data or both",
+                    default="all", choices=["all", "obs", "sample"])
+parser.add_argument("--version", help="which version to use", default="latest")
+parser.add_argument("--server", help="which server to use",
+                    default="https://pcmdiweb.llnl.gov/pss/pmpdata")
+parser.add_argument("--version_in_path", action="store_true", default=False,
+                    help="Append version in root path, avoids clobbering versions")
+parser.add_argument(
+    "--output-path", help="directory where to download", default=None)
+# parser.use("num_workers")
+p = parser.get_parameter()
+
+# Step1 prepare the paths to get the sample datafiles
+pth = tempfile.mkdtemp()
+files = []
+if p.dataset in ["all", "obs"]:  # ok we need obs
+    download_file(p.server, "obs_{}.txt".format(p.version), "obs.txt")
+    files.append("obs.txt")
+if p.dataset in ["all", "sample"]:
+    download_file(p.server, "sample_{}.txt".format(p.version), "sample.txt")
+    files.append("sample.txt")
+
+# Ok now we can download
+for file in files:
+    # First do we clobber or not?
+    pathout = p.output_path
+    if p.version_in_path:
+        with open(file) as f:
+            header = f.readline().strip()
+            version = header.split("_")[-1]
+            pathout = os.path.join(p.output_path, version)
+    cdat_info.download_sample_data_files(file, path=pathout)