diff --git a/xopt/generators/bayesian/bayesian_generator.py b/xopt/generators/bayesian/bayesian_generator.py
index 2d29a754..f9ced1f4 100644
--- a/xopt/generators/bayesian/bayesian_generator.py
+++ b/xopt/generators/bayesian/bayesian_generator.py
@@ -411,9 +411,12 @@ def generate(self, n_candidates: int):
             # dict to track runtimes
             timing_results = {}
 
+            training_data = self.get_training_data(self.data)
+            self._validate_contextual_variables_no_nan(training_data)
+
             # update internal model with internal data
             start_time = time.perf_counter()
-            model = self.train_model(self.get_training_data(self.data))
+            model = self.train_model(training_data)
             timing_results["training"] = time.perf_counter() - start_time
 
             # propose candidates given model
@@ -935,6 +938,20 @@ def get_model_input_bounds(self, data: pd.DataFrame) -> Dict[str, List[float]]:
 
         return variable_bounds
 
+    def _validate_contextual_variables_no_nan(self, data: pd.DataFrame):
+        """check to make sure that the last row of data does not contain NaN in any of the contextual variable columns"""
+        last_row = data.iloc[-1]
+        contextual_with_nan = [
+            name
+            for name in self.contextual_variables
+            if name in data.columns and pd.isna(last_row[name])
+        ]
+        if contextual_with_nan:
+            raise ValueError(
+                "latest row contains NaN in contextual variable columns: "
+                + ", ".join(contextual_with_nan)
+            )
+
     @property
     def _candidate_names(self):
         """variable names corresponding to generated candidates"""
diff --git a/xopt/tests/generators/bayesian/test_contextual_bo.py b/xopt/tests/generators/bayesian/test_contextual_bo.py
index ba98f994..8fae5297 100644
--- a/xopt/tests/generators/bayesian/test_contextual_bo.py
+++ b/xopt/tests/generators/bayesian/test_contextual_bo.py
@@ -137,6 +137,23 @@ def test_contextual_variable_bounds_from_data(self):
         assert np.isclose(bounds["x2"][0], 0.3 - padding)
         assert np.isclose(bounds["x2"][1], 0.7 + padding)
 
+    def test_contextual_variable_nan_in_last_row_raises(self):
+        generator = UpperConfidenceBoundGenerator(vocs=self.vocs)
+        data = pd.DataFrame(
+            {
+                "x1": np.linspace(0.0, 1.0, 5),
+                "x2": np.array([0.3, 0.4, 0.5, 0.6, np.nan]),
+                "y": np.ones(5),
+            }
+        )
+        generator.add_data(data)
+
+        with pytest.raises(
+            ValueError,
+            match="latest row contains NaN in contextual variable columns",
+        ):
+            generator.generate(1)
+
     def test_contextual_variable_explicit_domain_overrides_data(self):
         vocs = VOCS(
             variables={"x1": [0, 1], "x2": ContextualVariable(domain=[0.2, 0.4])},