diff --git a/src/c_api/c_api.cc b/src/c_api/c_api.cc
index 7cb7c15758cd..442304f653f9 100644
--- a/src/c_api/c_api.cc
+++ b/src/c_api/c_api.cc
@@ -198,11 +198,6 @@ void CreateNDArray(const DataType* shape,
                    int dtype,
                    NDArrayHandle* out) {
   mxnet::TShape requested_shape = mxnet::TShape(shape, shape + ndim);
-  if (!features::is_enabled(features::INT64_TENSOR_SIZE)) {
-    CHECK_LT(requested_shape.Size(), (int64_t{1} << 31) - 1) <<
-              "[CreateNDArray] Size of tensor you are trying to allocate is larger than "
-              "2^31 elements. Please build with flag USE_INT64_TENSOR_SIZE=1";
-  }
   *out = new NDArray(requested_shape,
                      Context::Create(static_cast<Context::DeviceType>(dev_type), dev_id),
                      delay_alloc != 0, dtype);
diff --git a/src/c_api/c_api_ndarray.cc b/src/c_api/c_api_ndarray.cc
index 6bfb3b35743d..c8b36d685c96 100644
--- a/src/c_api/c_api_ndarray.cc
+++ b/src/c_api/c_api_ndarray.cc
@@ -55,11 +55,6 @@ void SetNDInputsOutputs(const nnvm::Op* op,
   ndinputs->reserve(num_inputs);
   for (int i = 0; i < num_inputs; ++i) {
     NDArray* inp = reinterpret_cast<NDArray*>(inputs[i]);
-    if (!features::is_enabled(features::INT64_TENSOR_SIZE)) {
-      CHECK_LT(inp->shape().Size(), (int64_t{1} << 31) - 1) <<
-                "[SetNDInputsOutputs] Size of tensor you are trying to allocate is larger than "
-                "2^31 elements. Please build with flag USE_INT64_TENSOR_SIZE=1";
-    }
     ndinputs->emplace_back(inp);
   }
 
diff --git a/src/ndarray/ndarray.cc b/src/ndarray/ndarray.cc
index 14311a8795dc..aaa7aedf8bcd 100644
--- a/src/ndarray/ndarray.cc
+++ b/src/ndarray/ndarray.cc
@@ -142,11 +142,6 @@ void NDArray::Chunk::CheckAndAllocData(const mxnet::TShape &shape, int dtype) {
   CHECK_NE(aux_shapes.size(), 0)
       << "data is expected to be allocated after aux_data";
   auto dbytes = shape.Size() * mshadow::mshadow_sizeof(dtype);
-  if (!features::is_enabled(features::INT64_TENSOR_SIZE)) {
-    CHECK_LT(shape.Size(), (int64_t{1} << 31) - 1) <<
-              "[CheckAndAllocData] Size of tensor you are trying to allocate is larger than "
-              "2^31 elements. Please build with flag USE_INT64_TENSOR_SIZE=1";
-  }
   if (shandle.size < dbytes) {
     // free storage
     Storage::Get()->Free(shandle);
@@ -1887,11 +1882,6 @@ NDArray NDArray::Copy(Context ctx) const {
 
 void NDArray::SyncCopyFromCPU(const void *data, size_t size) const {
   mxnet::TShape dshape = this->shape();
-  if (!features::is_enabled(features::INT64_TENSOR_SIZE)) {
-    CHECK_LT(size, (int64_t{1} << 31) - 1) <<
-              "[SyncCopyFromCPU] Size of tensor you are trying to allocate is larger than "
-              "2^31 elements. Please build with flag USE_INT64_TENSOR_SIZE=1";
-  }
   CHECK_EQ(dshape.Size(), size)
       << "Memory size do not match";
   // zero-size array, no need to copy
@@ -2027,11 +2017,6 @@ void NDArray::SyncCopyFromNDArray(const NDArray& src, int i, int j) {
 
 void NDArray::SyncCopyToCPU(void *data, size_t size) const {
   mxnet::TShape dshape = this->shape();
-  if (!features::is_enabled(features::INT64_TENSOR_SIZE)) {
-    CHECK_LT(size, (int64_t{1} << 31) - 1) <<
-              "[SyncCopyToCPU] Size of tensor you are trying to allocate is larger than "
-              "2^31 elements. Please build with flag USE_INT64_TENSOR_SIZE=1";
-  }
   CHECK_EQ(dshape.Size(), size)
       << "Memory size do not match";
   // zero-size array, no need to copy
diff --git a/src/ndarray/ndarray_function.cc b/src/ndarray/ndarray_function.cc
index ed121899436a..34429446bd62 100644
--- a/src/ndarray/ndarray_function.cc
+++ b/src/ndarray/ndarray_function.cc
@@ -38,11 +38,6 @@ void Copy<cpu, cpu>(const TBlob &from, TBlob *to,
                     RunContext ctx) {
   MSHADOW_TYPE_SWITCH_WITH_BOOL(to->type_flag_, DType, {
     if (to->type_flag_ == from.type_flag_) {
-      if (!features::is_enabled(features::INT64_TENSOR_SIZE)) {
-        CHECK_LT(from.Size(), (int64_t{1} << 31) - 1) <<
-                  "Size of tensor you are trying to allocate is larger than "
-                  "2^31 elements. Please build with flag USE_INT64_TENSOR_SIZE=1";
-      }
       const index_t size = static_cast<index_t>(from.Size());
       CHECK_EQ(size, to->Size()) << "copying size mismatch, from: " << size * sizeof(DType)
                << " bytes, to: " << to->Size() * sizeof(DType) << " bytes.";
diff --git a/src/operator/tensor/init_op.h b/src/operator/tensor/init_op.h
index ac7be8366155..30ad1ded90c0 100644
--- a/src/operator/tensor/init_op.h
+++ b/src/operator/tensor/init_op.h
@@ -348,11 +348,6 @@ inline bool InitStorageType(const nnvm::NodeAttrs& attrs,
 template <bool is_integer = false, typename ValueType, typename xpu>
 void Fill(mshadow::Stream<xpu> *s, const TBlob& b, const OpReqType req, ValueType val) {
   // If b is a zero-size tensor, do nothing.
-  if (!features::is_enabled(features::INT64_TENSOR_SIZE)) {
-    CHECK_LT(b.Size(), (int64_t{1} << 31) - 1) <<
-              "[Fill] Size of tensor you are trying to allocate is larger than "
-              "2^31 elements. Please build with flag USE_INT64_TENSOR_SIZE=1";
-  }
   if (b.Size() == 0) return;
   if (req != kNullOp) {
     const size_t size = b.Size();
@@ -646,11 +641,6 @@ inline bool LinspaceShape(const nnvm::NodeAttrs& attrs,
   CHECK_GE(param.num, 0)
     << "Number of sequence should be non-negative, received " << param.num;
   mxnet::TShape shape = mxnet::TShape({static_cast<nnvm::dim_t>(param.num)});
-  if (!features::is_enabled(features::INT64_TENSOR_SIZE)) {
-    CHECK_LT(shape.Size(), (int64_t{1} << 31) - 1) <<
-              "[LinspaceShape] Size of tensor you are trying to allocate is larger than "
-              "2^32 elements. Please build with flag USE_INT64_TENSOR_SIZE=1";
-  }
   SHAPE_ASSIGN_CHECK(*out_attrs, 0, shape);
   return true;
 }