diff --git a/src/core/diags/README.md b/src/core/diags/README.md
index b5f4396ad..810069805 100644
--- a/src/core/diags/README.md
+++ b/src/core/diags/README.md
@@ -14,7 +14,7 @@ The diagnostics module supports common diagnostics features that are listed belo
 - [diag stream](#diag-stream-start)
 - [diag power](#diag-power)
 - [diag powersettings](#diag-powersettings)
-- [diag send](#diag-send-packets-length)
+- [diag send](#diag-send-async-packets-length)
 - [diag repeat](#diag-repeat-delay-length)
 - [diag radio](#diag-radio-sleep)
 - [diag rawpowersetting](#diag-rawpowersetting)
@@ -163,10 +163,11 @@ RawPowerSetting: 223344
 Done
 ```
 
-### diag send \<packets\> [length]
+### diag send [async] \<packets\> [length]
 
 Transmit a fixed number of packets.
 
+- async: Use the non-blocking mode.
 - packets: The number of packets to be sent.
 - length: The length of packet. The valid range is [3, 127].
 
@@ -175,6 +176,8 @@ Send the frame set by `diag frame` if length is omitted. Otherwise overwrite the
 ```bash
 > diag send 20 100
 Done
+> diag send async 20 100
+Done
 ```
 
 ### diag repeat \<delay\> [length]
diff --git a/src/core/diags/factory_diags.cpp b/src/core/diags/factory_diags.cpp
index de78c4d6d..204a80b4e 100644
--- a/src/core/diags/factory_diags.cpp
+++ b/src/core/diags/factory_diags.cpp
@@ -199,6 +199,7 @@ Diags::Diags(Instance &aInstance)
     , mTxPower(0)
     , mTxLen(0)
     , mIsTxPacketSet(false)
+    , mIsAsyncSend(false)
     , mRepeatActive(false)
     , mDiagSendOn(false)
     , mOutputCallback(nullptr)
@@ -429,6 +430,18 @@ Error Diags::ProcessSend(uint8_t aArgsLength, char *aArgs[])
 
     VerifyOrExit(aArgsLength >= 1, error = kErrorInvalidArgs);
 
+    if (StringMatch(aArgs[0], "async"))
+    {
+        aArgs++;
+        aArgsLength--;
+        VerifyOrExit(aArgsLength >= 1, error = kErrorInvalidArgs);
+        mIsAsyncSend = true;
+    }
+    else
+    {
+        mIsAsyncSend = false;
+    }
+
     SuccessOrExit(error = Utils::CmdLineParser::ParseAsUint32(aArgs[0], txPackets));
     mTxPackets = txPackets;
 
@@ -452,6 +465,11 @@ Error Diags::ProcessSend(uint8_t aArgsLength, char *aArgs[])
 
     TransmitPacket();
 
+    if (!mIsAsyncSend)
+    {
+        error = kErrorPending;
+    }
+
 exit:
     return error;
 }
@@ -852,11 +870,22 @@ void Diags::TransmitDone(Error aError)
         break;
     }
 
-    VerifyOrExit(!mRepeatActive);
-    VerifyOrExit(mTxPackets > 1);
-    mTxPackets--;
+    VerifyOrExit(!mRepeatActive && (mTxPackets > 0));
 
-    TransmitPacket();
+    if (mTxPackets > 1)
+    {
+        mTxPackets--;
+        TransmitPacket();
+    }
+    else
+    {
+        mTxPackets = 0;
+
+        if (!mIsAsyncSend)
+        {
+            Output("OT_ERROR_NONE");
+        }
+    }
 
 exit:
     return;
diff --git a/src/core/diags/factory_diags.hpp b/src/core/diags/factory_diags.hpp
index 82e68b71f..f06cafeeb 100644
--- a/src/core/diags/factory_diags.hpp
+++ b/src/core/diags/factory_diags.hpp
@@ -255,6 +255,7 @@ private:
     uint8_t       mTxLen;
     bool          mIsHeaderUpdated : 1;
     bool          mIsTxPacketSet : 1;
+    bool          mIsAsyncSend : 1;
     bool          mRepeatActive : 1;
     bool          mDiagSendOn : 1;
 #endif
diff --git a/tests/scripts/expect/cli-diags.exp b/tests/scripts/expect/cli-diags.exp
index e0f81e6d9..f7c8be8f2 100755
--- a/tests/scripts/expect/cli-diags.exp
+++ b/tests/scripts/expect/cli-diags.exp
@@ -57,8 +57,6 @@ expect_line "Done"
 send "diag send 10 100\n"
 expect_line "Done"
 
-sleep 2
-
 send "diag stats\n"
 expect "received packets: 0"
 expect "sent success packets: 10"
@@ -69,10 +67,25 @@ expect "first received packet: rssi=0, lqi=0"
 expect "last received packet: rssi=0, lqi=0"
 expect_line "Done"
 
+send "diag send async 10 100\n"
+expect_line "Done"
+
+sleep 2
+
+send "diag stats\n"
+expect "received packets: 0"
+expect "sent success packets: 20"
+expect "sent error cca packets: 0"
+expect "sent error abort packets: 0"
+expect "sent error others packets: 0"
+expect "first received packet: rssi=0, lqi=0"
+expect "last received packet: rssi=0, lqi=0"
+expect_line "Done"
+
 switch_node 1
 
 send "diag stats\n"
-expect "received packets: 10"
+expect "received packets: 20"
 expect "sent success packets: 0"
 expect "sent error cca packets: 0"
 expect "sent error abort packets: 0"
@@ -139,6 +152,9 @@ expect "Done"
 send "diag repeat 1\n"
 expect "Done"
 
+send "diag repeat stop\n"
+expect "Done"
+
 send_user "send frame with security processed\n"
 send "diag frame -s 112233\n"
 expect "Done"
diff --git a/tools/otci/otci/otci.py b/tools/otci/otci/otci.py
index 6f8aa964c..948919f72 100644
--- a/tools/otci/otci/otci.py
+++ b/tools/otci/otci/otci.py
@@ -2579,12 +2579,13 @@ class OTCI(object):
         """Stop transmitting a stream of characters."""
         self.execute_command('diag stream stop')
 
-    def diag_send(self, packets: int, length: Optional[int] = None):
+    def diag_send(self, packets: int, length: Optional[int] = None, is_async: bool = True):
         """Transmit a fixed number of packets."""
-        if length is None:
-            command = f'diag send {packets}'
-        else:
-            command = f'diag send {packets} {length}'
+        command = 'diag send '
+        command += 'async ' if is_async else ''
+        command += f'{packets} '
+        command += f'{length}' if length is not None else ''
+
         self.execute_command(command)
 
     def diag_repeat(self, delay: int, length: Optional[int] = None):