diff --git a/experiments/decision_tree/basic_dTree.py b/experiments/decision_tree/basic_dTree.py
new file mode 100644
index 0000000..08d950e
--- /dev/null
+++ b/experiments/decision_tree/basic_dTree.py
@@ -0,0 +1,110 @@
+import numpy as np
+
+'''
+Purpose:provide basic implementation of decision tree.
+'''
+
+class Leaf:
+    def __init__(self, y):
+        self.value = self._majority_class(y)
+
+    def _majority_class(self, y):
+        from collections import Counter
+        return Counter(y).most_common(1)[0][0]
+
+class Node:
+    def __init__(self, best_feature, best_threshold, left_child, right_child):
+        self.best_feature = best_feature
+        self.best_threshold = best_threshold
+        self.left_child = left_child
+        self.right_child = right_child
+
+class DecisionTree:
+    def __init__(self):
+        self.depth = 0
+
+    def build_tree(self, X, y, depth, max_depth):
+        if depth >= max_depth or len(set(y)) == 1:
+            return Leaf(y)
+
+        best_feature, best_threshold = self.find_best_split(X, y)
+
+        if best_feature is None:
+            return Leaf(y)
+
+        left_data, right_data, y_left, y_right = self.split(X, y, best_feature, best_threshold)
+
+        left_child = self.build_tree(left_data, y_left, depth + 1, max_depth)
+        right_child = self.build_tree(right_data, y_right, depth + 1, max_depth)
+
+        return Node(best_feature, best_threshold, left_child, right_child)
+    
+    def split(self, X, y, b_f, b_th):
+        left_data = []
+        right_data = []
+
+        y_left = []
+        y_right = []
+
+        for i in range(len(X)):
+            if X[i][b_f] <= b_th:
+                left_data.append(X[i])
+                y_left.append(y[i])
+            else:
+                right_data.append(X[i])
+                y_right.append(y[i])
+
+        return left_data, right_data, y_left, y_right
+
+    def calculate_gini(self, labels):
+        _, counts = np.unique(labels, return_counts=True)
+        probs = counts / counts.sum()
+        return 1 - np.sum(probs ** 2)
+
+    def find_best_split(self, X, y):
+        n_samples, n_features = len(X), len(X[0])
+        gini_parent = self.calculate_gini(y)
+
+        best_IG = 0
+        best_feature = None
+        best_threshold = None
+
+        X = np.array(X)  # ensure numpy
+        y = np.array(y)
+
+        for f in range(n_features):
+            values = np.unique(X[:, f])
+            thresholds = (values[:-1] + values[1:]) / 2
+
+            for t in thresholds:
+                left_idx = X[:, f] < t
+                right_idx = X[:, f] >= t
+
+                if len(y[left_idx]) == 0 or len(y[right_idx]) == 0:
+                    continue
+
+                gini_left = self.calculate_gini(y[left_idx])
+                gini_right = self.calculate_gini(y[right_idx])
+
+                curr_IG = gini_parent - (
+                    (len(y[left_idx]) / n_samples) * gini_left +
+                    (len(y[right_idx]) / n_samples) * gini_right
+                )
+
+                if curr_IG > best_IG:
+                    best_IG = curr_IG
+                    best_feature = f
+                    best_threshold = t
+
+        return best_feature, best_threshold
+    
+    def predict_sample(self, x, node):
+        # if leaf → return prediction
+        if isinstance(node, Leaf):
+            return node.value
+
+        # otherwise follow the split
+        if x[node.best_feature] <= node.best_threshold:
+            return self.predict_sample(x, node.left_child)
+        else:
+            return self.predict_sample(x, node.right_child)
\ No newline at end of file
diff --git a/experiments/decision_tree/run_basic_dtree.py b/experiments/decision_tree/run_basic_dtree.py
new file mode 100644
index 0000000..98bffcf
--- /dev/null
+++ b/experiments/decision_tree/run_basic_dtree.py
@@ -0,0 +1,25 @@
+from experiments.decision_tree.basic_dTree import DecisionTree
+
+'''
+Purpose: Test the basic implementation of decision tree
+'''
+
+X1 = [[1], [2], [3], [10], [11], [12]]
+y1 = [0, 0, 0, 1, 1, 1]
+
+X2 = [
+    [1, 1], [2, 1], [1, 2],
+    [8, 8], [9, 8], [8, 9]
+]
+y2 = [0, 0, 0, 1, 1, 1]
+
+X3 = [[1], [2], [3], [4], [5]]
+y3 = [0, 0, 1, 0, 1]
+
+model = DecisionTree()
+
+node = model.build_tree(X1, y1, 0, 3)
+
+pred = model.predict_sample([[6]], node)
+
+print(pred)
\ No newline at end of file
diff --git a/glassboxml/models/decision_tree.py b/glassboxml/models/decision_tree.py
new file mode 100644
index 0000000..dc9c80f
--- /dev/null
+++ b/glassboxml/models/decision_tree.py
@@ -0,0 +1,106 @@
+import numpy as np
+
+class Leaf:
+    def __init__(self, y):
+        self.value = self._majority_class(y)
+
+    def _majority_class(self, y):
+        from collections import Counter
+        return Counter(y).most_common(1)[0][0]
+    
+
+class Node:
+    def __init__(self, best_feature, best_threshold, left_child, right_child):
+        self.best_feature = best_feature
+        self.best_threshold = best_threshold
+        self.left_child = left_child
+        self.right_child = right_child
+
+class DecisionTree:
+    def __init__(self, max_depth):
+        self.max_depth = max_depth
+        self.root = None
+
+    def fit(self, X, y):
+        X = np.array(X)
+        y = np.array(y)
+
+        self.root = self.build_tree(X,y, 0)
+    
+    def build_tree(self, X, y, depth):
+        # if we either reach maximum depth or all of the elements left are of the same class, then return the leaf
+        if depth >= self.max_depth or len(set(y))==1:
+            return Leaf(y)
+        
+        best_feature, best_threshold = self.find_best_split(X, y)
+
+        left_data, right_data, y_left, y_right = self.split(X, y, best_feature, best_threshold)
+
+        left_child = self.build_tree(left_data, y_left, depth+1)
+        right_child = self.build_tree(right_data, y_right, depth+1)
+
+        return Node(best_feature, best_threshold, left_child, right_child)
+
+    def split(self, X, y, b_f, b_th):
+        left_data = []
+        right_data = []
+
+        y_left = []
+        y_right = []
+
+        for i in range(len(X)):
+            if X[i][b_f] <= b_th:
+                left_data.append(X[i])
+                y_left.append(y[i])
+            else:
+                right_data.append(X[i])
+                y_right.append(y[i])
+
+        return left_data, right_data, y_left, y_right
+
+    def calculate_gini(self, y):
+        _, counts = np.unique(y, return_counts=True)
+        probs = counts / counts.sum()
+        return 1 - np.sum(probs ** 2)
+
+    def find_best_split(self, X, y):
+        n_samples, n_features = X.shape
+        best_IG = 0
+        best_feature = 0
+        best_threshold = 0
+
+        gini_parent = self.calculate_gini(y)
+
+        for f in range(n_features):
+            values = np.unique(X[:, f])
+            thresholds = (values[:-1] + values[1:])/2
+            for t in thresholds:
+                left_idx = X[:,f] < t
+                right_idx = X[:,f] > t
+
+                if (len(y[left_idx]) == 0 or len(y[right_idx]) == 0):
+                    continue
+
+                gini_left = self.calculate_gini(y[left_idx])
+                gini_right = self.calculate_gini(y[right_idx])
+
+                curr_IG = gini_parent - (
+                    (len(y[left_idx]) / n_samples) * gini_left +
+                    (len(y[right_idx]) / n_samples) * gini_right
+                )
+
+                if curr_IG > best_IG:
+                    best_IG = curr_IG
+                    best_feature = f
+                    best_threshold = t
+        
+        return best_feature, best_threshold
+    
+    def predict_sample(self, x, node):
+        if isinstance(node, Leaf):
+            return node.value
+        
+        if x[node.best_feature] <= node.best_threshold:
+            return self.predict_sample(x, node.left_child)
+        else:
+            return self.predict_sample(x, node.right_child)
\ No newline at end of file
diff --git a/tests/models/test_decision_tree.py b/tests/models/test_decision_tree.py
new file mode 100644
index 0000000..465f938
--- /dev/null
+++ b/tests/models/test_decision_tree.py
@@ -0,0 +1,52 @@
+import numpy as np
+from glassboxml.models.decision_tree import DecisionTree, Leaf, Node
+
+def test_tree_perfect_split():
+    X = [[1], [2], [3], [10], [11], [12]]
+    y = [0, 0, 0, 1, 1, 1]
+
+    model = DecisionTree(max_depth=2)
+    model.fit(X, y)
+
+    preds = [model.predict_sample(x, model.root) for x in X]
+
+    assert preds == y, f"Predictions {preds} do not match labels {y}"
+
+def test_pure_node_returns_leaf():
+    X = [[1], [2], [3]]
+    y = [1, 1, 1]
+
+    model = DecisionTree(max_depth=3)
+    model.fit(X, y)
+
+    assert isinstance(model.root, Leaf)
+    assert model.root.value == 1
+
+def test_max_depth_limits_growth():
+    X = [[1], [2], [3], [10], [11], [12]]
+    y = [0, 0, 0, 1, 1, 1]
+
+    model = DecisionTree(max_depth=0)
+    model.fit(X, y)
+
+    assert isinstance(model.root, Leaf)
+
+def test_gini_impurity():
+    model = DecisionTree(max_depth=1)
+
+    y = np.array([0, 0, 1, 1])
+    gini = model.calculate_gini(y)
+
+    assert np.isclose(gini, 0.5), f"Gini {gini} should be 0.5"
+
+def test_predict_consistency():
+    X = [[1], [2], [3], [10], [11], [12]]
+    y = [0, 0, 0, 1, 1, 1]
+
+    model = DecisionTree(max_depth=2)
+    model.fit(X, y)
+
+    preds1 = [model.predict_sample(x, model.root) for x in X]
+    preds2 = [model.predict_sample(x, model.root) for x in X]
+
+    assert preds1 == preds2
\ No newline at end of file