diff --git a/test/test_tensor.py b/test/test_tensor.py
index af76fe0b67..5c745a6307 100644
--- a/test/test_tensor.py
+++ b/test/test_tensor.py
@@ -145,6 +145,13 @@ class TestTinygrad(unittest.TestCase):
         b = random_fn(10,10).realize()
         np.testing.assert_allclose(a.numpy(), b.numpy())
 
+  def test_randn_isnt_inf_on_zero(self):
+    # simulate failure case of rand handing a zero to randn
+    original_rand, Tensor.rand = Tensor.rand, Tensor.zeros
+    try: self.assertNotIn(np.inf, Tensor.randn(16).numpy())
+    except: raise
+    finally: Tensor.rand = original_rand
+
   def test_zeros_like_has_same_dtype(self):
     for datatype in [dtypes.float16, dtypes.float32, dtypes.int8, dtypes.int32, dtypes.int64, dtypes.uint8]:
       a = Tensor([1, 2, 3], dtype=datatype)
diff --git a/tinygrad/tensor.py b/tinygrad/tensor.py
index d05cb7fb9f..75f67279c7 100644
--- a/tinygrad/tensor.py
+++ b/tinygrad/tensor.py
@@ -177,7 +177,7 @@ class Tensor:
   def randn(*shape, dtype:Optional[DType]=None, **kwargs) -> Tensor:
     # https://en.wikipedia.org/wiki/Box%E2%80%93Muller_transform
     src = Tensor.rand(2, *shape, **kwargs)
-    return src[0].mul(2*pi).cos().mul(src[1].log().mul(-2).sqrt()).cast(Tensor.default_type if dtype is None else dtype)
+    return src[0].mul(2*pi).cos().mul(((1 - src[1])).log().mul(-2).sqrt()).cast(Tensor.default_type if dtype is None else dtype)
 
   @staticmethod
   def uniform(*shape, low=-1.0, high=1.0, **kwargs) -> Tensor: return ((high-low) * Tensor.rand(*shape, **kwargs)) + low