diff --git a/audio/channel_send.cc b/audio/channel_send.cc
index d4266c2bc5..e8360cb4a1 100644
--- a/audio/channel_send.cc
+++ b/audio/channel_send.cc
@@ -495,15 +495,6 @@ int32_t ChannelSend::SendRtpAudio(AudioFrameType frameType,
                                   uint8_t payloadType,
                                   uint32_t timeStamp,
                                   rtc::ArrayView<const uint8_t> payload) {
-  if (payload.empty()) {
-    if (frameType == AudioFrameType::kEmptyFrame) {
-      // we don't send empty audio RTP packets
-      // no error since we use it to drive DTMF when we use VAD
-      return 0;
-    }
-    return -1;
-  }
-
   if (_includeAudioLevelIndication) {
     // Store current audio level in the RTP sender.
     // The level will be used in combination with voice-activity state
diff --git a/modules/rtp_rtcp/source/rtp_sender_audio.cc b/modules/rtp_rtcp/source/rtp_sender_audio.cc
index 168f2061f4..86026ffd92 100644
--- a/modules/rtp_rtcp/source/rtp_sender_audio.cc
+++ b/modules/rtp_rtcp/source/rtp_sender_audio.cc
@@ -221,6 +221,14 @@ bool RTPSenderAudio::SendAudio(AudioFrameType frame_type,
     }
     return true;
   }
+  if (payload_size == 0 || payload_data == NULL) {
+    if (frame_type == AudioFrameType::kEmptyFrame) {
+      // we don't send empty audio RTP packets
+      // no error since we use it to drive DTMF when we use VAD
+      return true;
+    }
+    return false;
+  }
 
   std::unique_ptr<RtpPacketToSend> packet = rtp_sender_->AllocatePacket();
   packet->SetMarker(MarkerBit(frame_type, payload_type));