diff --git a/pkg/tcpip/transport/tcp/connect.go b/pkg/tcpip/transport/tcp/connect.go
index 0dc710276..a00ef97c6 100644
--- a/pkg/tcpip/transport/tcp/connect.go
+++ b/pkg/tcpip/transport/tcp/connect.go
@@ -1357,6 +1357,7 @@ func (e *endpoint) protocolMainLoop(handshake bool, wakerInitDone chan<- struct{
 		// e.mu is expected to be hold upon entering this section.
 		if e.snd != nil {
 			e.snd.resendTimer.cleanup()
+			e.snd.rc.probeTimer.cleanup()
 		}
 
 		if closeTimer != nil {
@@ -1436,6 +1437,10 @@ func (e *endpoint) protocolMainLoop(handshake bool, wakerInitDone chan<- struct{
 				return nil
 			},
 		},
+		{
+			w: &e.snd.rc.probeWaker,
+			f: e.snd.probeTimerExpired,
+		},
 		{
 			w: &e.newSegmentWaker,
 			f: func() *tcpip.Error {
diff --git a/pkg/tcpip/transport/tcp/endpoint.go b/pkg/tcpip/transport/tcp/endpoint.go
index 8f3981075..281f4cd58 100644
--- a/pkg/tcpip/transport/tcp/endpoint.go
+++ b/pkg/tcpip/transport/tcp/endpoint.go
@@ -508,6 +508,9 @@ type endpoint struct {
 	// shutdownFlags represent the current shutdown state of the endpoint.
 	shutdownFlags tcpip.ShutdownFlags
 
+	// tcpRecovery is the loss deteoction algorithm used by TCP.
+	tcpRecovery tcpip.TCPRecovery
+
 	// sackPermitted is set to true if the peer sends the TCPSACKPermitted
 	// option in the SYN/SYN-ACK.
 	sackPermitted bool
@@ -918,6 +921,8 @@ func newEndpoint(s *stack.Stack, netProto tcpip.NetworkProtocolNumber, waiterQue
 		e.maxSynRetries = uint8(synRetries)
 	}
 
+	s.TransportProtocolOption(ProtocolNumber, &e.tcpRecovery)
+
 	if p := s.GetTCPProbe(); p != nil {
 		e.probe = p
 	}
@@ -3072,7 +3077,7 @@ func (e *endpoint) completeState() stack.TCPEndpointState {
 		}
 	}
 
-	rc := e.snd.rc
+	rc := &e.snd.rc
 	s.Sender.RACKState = stack.TCPRACKState{
 		XmitTime:    rc.xmitTime,
 		EndSequence: rc.endSequence,
diff --git a/pkg/tcpip/transport/tcp/protocol.go b/pkg/tcpip/transport/tcp/protocol.go
index 672159eed..c9e194f82 100644
--- a/pkg/tcpip/transport/tcp/protocol.go
+++ b/pkg/tcpip/transport/tcp/protocol.go
@@ -405,7 +405,7 @@ func (p *protocol) Option(option tcpip.GettableTransportProtocolOption) *tcpip.E
 
 	case *tcpip.TCPRecovery:
 		p.mu.RLock()
-		*v = tcpip.TCPRecovery(p.recovery)
+		*v = p.recovery
 		p.mu.RUnlock()
 		return nil
 
@@ -543,7 +543,8 @@ func NewProtocol(s *stack.Stack) stack.TransportProtocol {
 		minRTO:                     MinRTO,
 		maxRTO:                     MaxRTO,
 		maxRetries:                 MaxRetries,
-		recovery:                   tcpip.TCPRACKLossDetection,
+		// TODO(gvisor.dev/issue/5243): Set recovery to tcpip.TCPRACKLossDetection.
+		recovery: 0,
 	}
 	p.dispatcher.init(runtime.GOMAXPROCS(0))
 	return &p
diff --git a/pkg/tcpip/transport/tcp/rack.go b/pkg/tcpip/transport/tcp/rack.go
index e0a50a919..b71e6b992 100644
--- a/pkg/tcpip/transport/tcp/rack.go
+++ b/pkg/tcpip/transport/tcp/rack.go
@@ -17,9 +17,18 @@ package tcp
 import (
 	"time"
 
+	"gvisor.dev/gvisor/pkg/sleep"
+	"gvisor.dev/gvisor/pkg/tcpip"
 	"gvisor.dev/gvisor/pkg/tcpip/seqnum"
 )
 
+// wcDelayedACKTimeout is the recommended maximum delayed ACK timer value as
+// defined in https://tools.ietf.org/html/draft-ietf-tcpm-rack-08#section-7.5.
+// It stands for worst case delayed ACK timer (WCDelAckT). When FlightSize is
+// 1, PTO is inflated by WCDelAckT time to compensate for a potential long
+// delayed ACK timer at the receiver.
+const wcDelayedACKTimeout = 200 * time.Millisecond
+
 // RACK is a loss detection algorithm used in TCP to detect packet loss and
 // reordering using transmission timestamp of the packets instead of packet or
 // sequence counts. To use RACK, SACK should be enabled on the connection.
@@ -54,6 +63,15 @@ type rackControl struct {
 
 	// xmitTime is the latest transmission timestamp of rackControl.seg.
 	xmitTime time.Time `state:".(unixTime)"`
+
+	// probeTimer and probeWaker are used to schedule PTO for RACK TLP algorithm.
+	probeTimer timer       `state:"nosave"`
+	probeWaker sleep.Waker `state:"nosave"`
+}
+
+// init initializes RACK specific fields.
+func (rc *rackControl) init() {
+	rc.probeTimer.init(&rc.probeWaker)
 }
 
 // update will update the RACK related fields when an ACK has been received.
@@ -127,3 +145,61 @@ func (rc *rackControl) detectReorder(seg *segment) {
 func (rc *rackControl) setDSACKSeen() {
 	rc.dsackSeen = true
 }
+
+// shouldSchedulePTO dictates whether we should schedule a PTO or not.
+// See https://tools.ietf.org/html/draft-ietf-tcpm-rack-08#section-7.5.1.
+func (s *sender) shouldSchedulePTO() bool {
+	// Schedule PTO only if RACK loss detection is enabled.
+	return s.ep.tcpRecovery&tcpip.TCPRACKLossDetection != 0 &&
+		// The connection supports SACK.
+		s.ep.sackPermitted &&
+		// The connection is not in loss recovery.
+		(s.state != RTORecovery && s.state != SACKRecovery) &&
+		// The connection has no SACKed sequences in the SACK scoreboard.
+		s.ep.scoreboard.Sacked() == 0
+}
+
+// schedulePTO schedules the probe timeout as defined in
+// https://tools.ietf.org/html/draft-ietf-tcpm-rack-08#section-7.5.1.
+func (s *sender) schedulePTO() {
+	pto := time.Second
+	s.rtt.Lock()
+	if s.rtt.srttInited && s.rtt.srtt > 0 {
+		pto = s.rtt.srtt * 2
+		if s.outstanding == 1 {
+			pto += wcDelayedACKTimeout
+		}
+	}
+	s.rtt.Unlock()
+
+	now := time.Now()
+	if s.resendTimer.enabled() {
+		if now.Add(pto).After(s.resendTimer.target) {
+			pto = s.resendTimer.target.Sub(now)
+		}
+		s.resendTimer.disable()
+	}
+
+	s.rc.probeTimer.enable(pto)
+}
+
+// probeTimerExpired is the same as TLP_send_probe() as defined in
+// https://tools.ietf.org/html/draft-ietf-tcpm-rack-08#section-7.5.2.
+func (s *sender) probeTimerExpired() *tcpip.Error {
+	if !s.rc.probeTimer.checkExpiration() {
+		return nil
+	}
+	// TODO(gvisor.dev/issue/5084): Implement this pseudo algorithm.
+	// 	If an unsent segment exists AND
+	// 			the receive window allows new data to be sent:
+	// 					Transmit the lowest-sequence unsent segment of up to SMSS
+	// 					Increment FlightSize by the size of the newly-sent segment
+	// 	Else if TLPRxtOut is not set:
+	// 					Retransmit the highest-sequence segment sent so far
+	// 					TLPRxtOut = true
+	// 					TLPHighRxt = SND.NXT
+	// 	The cwnd remains unchanged
+	//  If FlightSize != 0:
+	//  				Arm RTO timer only.
+	return nil
+}
diff --git a/pkg/tcpip/transport/tcp/rack_state.go b/pkg/tcpip/transport/tcp/rack_state.go
index c9dc7e773..76cad0831 100644
--- a/pkg/tcpip/transport/tcp/rack_state.go
+++ b/pkg/tcpip/transport/tcp/rack_state.go
@@ -27,3 +27,8 @@ func (rc *rackControl) saveXmitTime() unixTime {
 func (rc *rackControl) loadXmitTime(unix unixTime) {
 	rc.xmitTime = time.Unix(unix.second, unix.nano)
 }
+
+// afterLoad is invoked by stateify.
+func (rc *rackControl) afterLoad() {
+	rc.probeTimer.init(&rc.probeWaker)
+}
diff --git a/pkg/tcpip/transport/tcp/snd.go b/pkg/tcpip/transport/tcp/snd.go
index cc991aba6..c0e9d98e3 100644
--- a/pkg/tcpip/transport/tcp/snd.go
+++ b/pkg/tcpip/transport/tcp/snd.go
@@ -286,6 +286,8 @@ func newSender(ep *endpoint, iss, irs seqnum.Value, sndWnd seqnum.Size, mss uint
 		gso: ep.gso != nil,
 	}
 
+	s.rc.init()
+
 	if s.gso {
 		s.ep.gso.MSS = uint16(maxPayloadSize)
 	}
@@ -1455,6 +1457,7 @@ func (s *sender) handleRcvdSegment(rcvdSeg *segment) {
 			// Reset firstRetransmittedSegXmitTime to the zero value.
 			s.firstRetransmittedSegXmitTime = time.Time{}
 			s.resendTimer.disable()
+			s.rc.probeTimer.disable()
 		}
 	}
 
diff --git a/test/syscalls/linux/proc_net.cc b/test/syscalls/linux/proc_net.cc
index 1cc700fe7..73140b2e9 100644
--- a/test/syscalls/linux/proc_net.cc
+++ b/test/syscalls/linux/proc_net.cc
@@ -499,7 +499,13 @@ TEST(ProcSysNetIpv4Recovery, CanReadAndWrite) {
   // Check initial value is set to 1.
   EXPECT_THAT(PreadFd(fd.get(), &buf, sizeof(buf), 0),
               SyscallSucceedsWithValue(sizeof(to_write) + 1));
-  EXPECT_EQ(strcmp(buf, "1\n"), 0);
+  if (IsRunningOnGvisor()) {
+    // TODO(gvisor.dev/issue/5243): TCPRACKLossDetection = 1 should be turned on
+    // by default.
+    EXPECT_EQ(strcmp(buf, "0\n"), 0);
+  } else {
+    EXPECT_EQ(strcmp(buf, "1\n"), 0);
+  }
 
   // Set tcp_recovery to one of the allowed constants.
   EXPECT_THAT(PwriteFd(fd.get(), &to_write, sizeof(to_write), 0),