diff --git a/datasets/psm.py b/datasets/psm.py
new file mode 100644
index 0000000..f812e2d
--- /dev/null
+++ b/datasets/psm.py
@@ -0,0 +1,62 @@
+from benchopt import BaseDataset, safe_import_context
+
+with safe_import_context() as import_ctx:
+    import os
+    import requests
+    import pandas as pd
+
+url_xtrain = (
+    "https://drive.google.com/uc?&id=1d3tAbYTj0CZLhB7z3IDTfTRg3E7qj_tw"
+    "&export=download"
+)
+url_xtest = (
+    "https://drive.google.com/uc?&id=1RQH7igHhm_0GAgXyVpkJk6TenDl9rd53"
+    "&export=download"
+)
+url_ytest = (
+    "https://drive.google.com/uc?&id=1SYgcRt0DH--byFbvkKTkezJKU5ZENZhw"
+    "&export=download"
+)
+
+
+class Dataset(BaseDataset):
+    name = "PSM"
+    install_cmd = "conda"
+    requirements = ["pandas"]
+    parameters = {
+        "debug": [False],
+    }
+
+    def get_data(self):
+        # Check if the data is already here
+        if not os.path.exists("data/PSM/PSM_train.csv"):
+            os.makedirs("data/PSM", exist_ok=True)
+            response = requests.get(url_xtrain)
+            with open("data/PSM/PSM_train.csv", "wb") as f:
+                f.write(response.content)
+            response = requests.get(url_xtest)
+            with open("data/PSM/PSM_test.csv", "wb") as f:
+                f.write(response.content)
+            response = requests.get(url_ytest)
+            with open("data/PSM/PSM_test_label.csv", "wb") as f:
+                f.write(response.content)
+
+        X_train = pd.read_csv("data/PSM/PSM_train.csv")
+        X_train.fillna(X_train.mean(), inplace=True)
+        X_train = X_train.to_numpy()
+
+        X_test = pd.read_csv("data/PSM/PSM_test.csv")
+        X_test.fillna(X_test.mean(), inplace=True)
+        X_test = X_test.to_numpy()
+
+        y_test = pd.read_csv("data/PSM/PSM_test_label.csv").to_numpy()[:, 1]
+
+        # Limiting the size of the dataset for testing purposes
+        if self.debug:
+            X_train = X_train[:1000]
+            X_test = X_test[:1000]
+            y_test = y_test[:1000]
+
+        return dict(
+            X_train=X_train, y_test=y_test, X_test=X_test
+        )