diff --git a/src/operator/tensor/broadcast_reduce_op.h b/src/operator/tensor/broadcast_reduce_op.h
index ea9aa6a7cf65..c4e3dae35ab9 100644
--- a/src/operator/tensor/broadcast_reduce_op.h
+++ b/src/operator/tensor/broadcast_reduce_op.h
@@ -1630,7 +1630,7 @@ template<int ndim, bool clip = true>
 struct pick {
   template<typename DType, typename IType>
   MSHADOW_XINLINE static void Map(index_t i, DType* out, const DType* a,
-                                  const IType *idx, index_t M, int stride,
+                                  const IType *idx, index_t M, index_t stride,
                                   mshadow::Shape<ndim> bshape,
                                   mshadow::Shape<ndim> sshape) {
     using namespace mxnet_op;
@@ -1652,7 +1652,7 @@ template<int ndim, bool clip = true>
 struct pick_grad {
   template<typename DType, typename IType>
   MSHADOW_XINLINE static void Map(index_t i, DType* igrad, const DType* ograd,
-                                  const IType *idx, index_t M, int stride,
+                                  const IType *idx, index_t M, index_t stride,
                                   mshadow::Shape<ndim> bshape,
                                   mshadow::Shape<ndim> sshape) {
     using namespace mxnet_op;
@@ -1717,7 +1717,7 @@ void PickOpForward(const nnvm::NodeAttrs& attrs,
 
   const mxnet::TShape& ishape = inputs[0].shape_;
   index_t axis = CheckAxis(param.axis.value(), ishape.ndim());
-  int leading = 1, trailing = 1, M = ishape[axis];
+  index_t leading = 1, trailing = 1, M = ishape[axis];
   for (index_t i = 0; i < axis; ++i) leading *= ishape[i];
   for (index_t i = axis+1; i < ishape.ndim(); ++i) trailing *= ishape[i];
 
@@ -1764,7 +1764,7 @@ void PickOpBackward(const nnvm::NodeAttrs& attrs,
 
   const mxnet::TShape& ishape = outputs[0].shape_;
   const index_t axis = CheckAxis(param.axis.value(), ishape.ndim());
-  int leading = 1, trailing = 1, M = ishape[axis];
+  index_t leading = 1, trailing = 1, M = ishape[axis];
   for (index_t i = 0; i < axis; ++i) leading *= ishape[i];
   for (index_t i = axis+1; i < ishape.ndim(); ++i) trailing *= ishape[i];
 
diff --git a/tests/nightly/test_np_large_array.py b/tests/nightly/test_np_large_array.py
index 90fc58f0f27b..d2b653b500cd 100644
--- a/tests/nightly/test_np_large_array.py
+++ b/tests/nightly/test_np_large_array.py
@@ -709,16 +709,18 @@ def test_one_hot():
     assert A.grad[0] == 0
 
 @use_np
-@pytest.mark.skip(reason='backward value broken on large tensor')
 def test_pick():
+    INT_OVERFLOW = 2**31
     A = np.zeros((INT_OVERFLOW, 2))
     B = np.zeros((INT_OVERFLOW))
+    A[-1, 0] = 3
     A.attach_grad()
     B.attach_grad()
     with mx.autograd.record():
         C = npx.pick(A, B)
     assert C.shape == (INT_OVERFLOW, )
     assert C[0] == 0
+    assert C[-1] == 3
     C.backward()
     assert A.grad.shape == (INT_OVERFLOW, 2)
     assert B.grad.shape == (INT_OVERFLOW, )